blob: a19b4ab3b7115efea9a80fff80d81b54b0a50e6b [file] [log] [blame]
<!DOCTYPE html>
<html>
<head>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width, initial-scale=1.0" /><meta name="generator" content="Docutils 0.17.1: http://docutils.sourceforge.net/" />
<title>Arrow Columnar Format &#8212; Apache Arrow v8.0.0</title>
<!-- Loaded before other Sphinx assets -->
<link href="../_static/styles/theme.css?digest=1999514e3f237ded88cf" rel="stylesheet">
<link href="../_static/styles/pydata-sphinx-theme.css?digest=1999514e3f237ded88cf" rel="stylesheet">
<link rel="stylesheet"
href="../_static/vendor/fontawesome/5.13.0/css/all.min.css">
<link rel="preload" as="font" type="font/woff2" crossorigin
href="../_static/vendor/fontawesome/5.13.0/webfonts/fa-solid-900.woff2">
<link rel="preload" as="font" type="font/woff2" crossorigin
href="../_static/vendor/fontawesome/5.13.0/webfonts/fa-brands-400.woff2">
<link rel="stylesheet" type="text/css" href="../_static/pygments.css" />
<link rel="stylesheet" type="text/css" href="../_static/copybutton.css" />
<link rel="stylesheet" type="text/css" href="../_static/design-style.b7bb847fb20b106c3d81b95245e65545.min.css" />
<link rel="stylesheet" type="text/css" href="../_static/theme_overrides.css" />
<!-- Pre-loaded scripts that we'll load fully later -->
<link rel="preload" as="script" href="../_static/scripts/pydata-sphinx-theme.js?digest=1999514e3f237ded88cf">
<script data-url_root="../" id="documentation_options" src="../_static/documentation_options.js"></script>
<script src="../_static/jquery.js"></script>
<script src="../_static/underscore.js"></script>
<script src="../_static/doctools.js"></script>
<script src="../_static/clipboard.min.js"></script>
<script src="../_static/copybutton.js"></script>
<script src="../_static/design-tabs.js"></script>
<link rel="canonical" href="https://arrow.apache.org/docs/format/Columnar.html" />
<link rel="shortcut icon" href="../_static/favicon.ico"/>
<link rel="index" title="Index" href="../genindex.html" />
<link rel="search" title="Search" href="../search.html" />
<link rel="next" title="Arrow Flight RPC" href="Flight.html" />
<link rel="prev" title="Format Versioning and Stability" href="Versioning.html" />
<meta name="viewport" content="width=device-width, initial-scale=1" />
<meta name="docsearch:language" content="None">
<!-- Google Analytics -->
<!-- Matomo -->
<script>
var _paq = window._paq = window._paq || [];
/* tracker methods like "setCustomDimension" should be called before "trackPageView" */
/* We explicitly disable cookie tracking to avoid privacy issues */
_paq.push(['disableCookies']);
_paq.push(['trackPageView']);
_paq.push(['enableLinkTracking']);
(function() {
var u="https://analytics.apache.org/";
_paq.push(['setTrackerUrl', u+'matomo.php']);
_paq.push(['setSiteId', '20']);
var d=document, g=d.createElement('script'), s=d.getElementsByTagName('script')[0];
g.async=true; g.src=u+'matomo.js'; s.parentNode.insertBefore(g,s);
})();
</script>
<!-- End Matomo Code -->
</head>
<body data-spy="scroll" data-target="#bd-toc-nav" data-offset="60">
<div class="container-fluid" id="banner"></div>
<div class="container-xl">
<div class="row">
<!-- Only show if we have sidebars configured, else just a small margin -->
<div class="col-12 col-md-3 bd-sidebar">
<div class="sidebar-start-items">
<a class="navbar-brand" href="../index.html">
<img src="../_static/arrow.png" class="logo" alt="logo">
</a>
<div id="version-search-wrapper">
<div id="version-button" class="dropdown">
<button type="button" class="btn btn-secondary btn-sm navbar-btn dropdown-toggle" id="version_switcher_button" data-toggle="dropdown">
8.0.0
<span class="caret"></span>
</button>
<div id="version_switcher" class="dropdown-menu list-group-flush py-0" aria-labelledby="version_switcher_button">
<!-- dropdown will be populated by javascript on page load -->
</div>
</div>
<script type="text/javascript">
// Function to construct the target URL from the JSON components
function buildURL(entry) {
var template = "https://arrow.apache.org/docs/{version}"; // supplied by jinja
template = template.replace("{version}", entry.version);
return template;
}
// Function to check if corresponding page path exists in other version of docs
// and, if so, go there instead of the homepage of the other docs version
function checkPageExistsAndRedirect(event) {
const currentFilePath = "format/Columnar.html",
otherDocsHomepage = event.target.getAttribute("href");
let tryUrl = `${otherDocsHomepage}${currentFilePath}`;
$.ajax({
type: 'HEAD',
url: tryUrl,
// if the page exists, go there
success: function() {
location.href = tryUrl;
}
}).fail(function() {
location.href = otherDocsHomepage;
});
return false;
}
// Function to populate the version switcher
(function () {
// get JSON config
$.getJSON("/docs/_static/versions.json", function(data, textStatus, jqXHR) {
// create the nodes first (before AJAX calls) to ensure the order is
// correct (for now, links will go to doc version homepage)
$.each(data, function(index, entry) {
// if no custom name specified (e.g., "latest"), use version string
if (!("name" in entry)) {
entry.name = entry.version;
}
// construct the appropriate URL, and add it to the dropdown
entry.url = buildURL(entry);
const node = document.createElement("a");
node.setAttribute("class", "list-group-item list-group-item-action py-1");
node.setAttribute("href", `${entry.url}`);
node.textContent = `${entry.name}`;
node.onclick = checkPageExistsAndRedirect;
$("#version_switcher").append(node);
});
});
})();
</script>
<form id="search-box" class="bd-search d-flex align-items-center" action="../search.html" method="get">
<i class="icon fas fa-search"></i>
<input type="search" class="form-control" name="q" id="search-input" placeholder="Search the docs ..." aria-label="Search the docs ..." autocomplete="off" >
</form>
</div>
<nav class="bd-links" id="bd-docs-nav" aria-label="Main navigation">
<div class="bd-toc-item active">
<p aria-level="2" class="caption" role="heading">
<span class="caption-text">
Supported Environments
</span>
</p>
<ul class="nav bd-sidenav">
<li class="toctree-l1">
<a class="reference internal" href="../c_glib/index.html">
C/GLib
</a>
</li>
<li class="toctree-l1 has-children">
<a class="reference internal" href="../cpp/index.html">
C++
</a>
<input class="toctree-checkbox" id="toctree-checkbox-1" name="toctree-checkbox-1" type="checkbox"/>
<label for="toctree-checkbox-1">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l2 has-children">
<a class="reference internal" href="../cpp/getting_started.html">
User Guide
</a>
<input class="toctree-checkbox" id="toctree-checkbox-2" name="toctree-checkbox-2" type="checkbox"/>
<label for="toctree-checkbox-2">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/overview.html">
High-Level Overview
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/conventions.html">
Conventions
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/build_system.html">
Using Arrow C++ in your own project
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/memory.html">
Memory Management
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/arrays.html">
Arrays
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/datatypes.html">
Data Types
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/tables.html">
Tabular Data
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/compute.html">
Compute Functions
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/streaming_execution.html">
Streaming execution engine
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/io.html">
Input / output and filesystems
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/ipc.html">
Reading and writing the Arrow IPC format
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/orc.html">
Reading and Writing ORC files
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/parquet.html">
Reading and writing Parquet files
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/csv.html">
Reading and Writing CSV files
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/json.html">
Reading JSON files
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/dataset.html">
Tabular Datasets
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/flight.html">
Arrow Flight RPC
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/gdb.html">
Debugging code using Arrow
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/threading.html">
Thread Management
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/env_vars.html">
Environment Variables
</a>
</li>
</ul>
</li>
<li class="toctree-l2 has-children">
<a class="reference internal" href="../cpp/examples/index.html">
Examples
</a>
<input class="toctree-checkbox" id="toctree-checkbox-3" name="toctree-checkbox-3" type="checkbox"/>
<label for="toctree-checkbox-3">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/examples/cmake_minimal_build.html">
Minimal build using CMake
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/examples/compute_and_write_example.html">
Compute and Write CSV Example
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/examples/dataset_documentation_example.html">
Arrow Datasets example
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/examples/row_columnar_conversion.html">
Row to columnar conversion
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/examples/tuple_range_conversion.html">
std::tuple-like ranges to Arrow
</a>
</li>
</ul>
</li>
<li class="toctree-l2 has-children">
<a class="reference internal" href="../cpp/api.html">
API Reference
</a>
<input class="toctree-checkbox" id="toctree-checkbox-4" name="toctree-checkbox-4" type="checkbox"/>
<label for="toctree-checkbox-4">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/support.html">
Programming Support
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/memory.html">
Memory (management)
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/thread.html">
Thread (management)
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/datatype.html">
Data Types
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/array.html">
Arrays
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/scalar.html">
Scalars
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/builder.html">
Array Builders
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/table.html">
Two-dimensional Datasets
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/c_abi.html">
C Interfaces
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/compute.html">
Compute Functions
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/tensor.html">
Tensors
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/utilities.html">
Utilities
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/async.html">
Asynchronous programming
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/io.html">
Input / output
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/ipc.html">
Arrow IPC
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/formats.html">
File Formats
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/cuda.html">
CUDA support
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/flight.html">
Arrow Flight RPC
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/flightsql.html">
Arrow Flight SQL
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/filesystem.html">
Filesystems
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../cpp/api/dataset.html">
Dataset
</a>
</li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l1">
<a class="reference external" href="https://github.com/apache/arrow/blob/master/csharp/README.md">
C#
</a>
</li>
<li class="toctree-l1">
<a class="reference external" href="https://pkg.go.dev/github.com/apache/arrow/go">
Go
</a>
</li>
<li class="toctree-l1 has-children">
<a class="reference internal" href="../java/index.html">
Java
</a>
<input class="toctree-checkbox" id="toctree-checkbox-5" name="toctree-checkbox-5" type="checkbox"/>
<label for="toctree-checkbox-5">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l2">
<a class="reference internal" href="../java/quickstartguide.html">
Quick Start Guide
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../java/overview.html">
High-Level Overview
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../java/install.html">
Installing Java Modules
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../java/memory.html">
Memory Management
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../java/vector.html">
ValueVector
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../java/vector_schema_root.html">
Tabular Data
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../java/ipc.html">
Reading/Writing IPC formats
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../java/algorithm.html">
Java Algorithms
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../java/flight.html">
Arrow Flight RPC
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../java/dataset.html">
Dataset
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../java/cdata.html">
C Data Interface
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../java/reference/index.html">
Reference (javadoc)
</a>
</li>
</ul>
</li>
<li class="toctree-l1">
<a class="reference internal" href="../js/index.html">
JavaScript
</a>
</li>
<li class="toctree-l1">
<a class="reference external" href="https://github.com/apache/arrow-julia/blob/main/README.md">
Julia
</a>
</li>
<li class="toctree-l1">
<a class="reference external" href="https://github.com/apache/arrow/blob/master/matlab/README.md">
MATLAB
</a>
</li>
<li class="toctree-l1 has-children">
<a class="reference internal" href="../python/index.html">
Python
</a>
<input class="toctree-checkbox" id="toctree-checkbox-6" name="toctree-checkbox-6" type="checkbox"/>
<label for="toctree-checkbox-6">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l2">
<a class="reference internal" href="../python/install.html">
Installing PyArrow
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/getstarted.html">
Getting Started
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/data.html">
Data Types and In-Memory Data Model
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/compute.html">
Compute Functions
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/memory.html">
Memory and IO Interfaces
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/ipc.html">
Streaming, Serialization, and IPC
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/filesystems.html">
Filesystem Interface
</a>
</li>
<li class="toctree-l2 has-children">
<a class="reference internal" href="../python/filesystems_deprecated.html">
Filesystem Interface (legacy)
</a>
<input class="toctree-checkbox" id="toctree-checkbox-7" name="toctree-checkbox-7" type="checkbox"/>
<label for="toctree-checkbox-7">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.hdfs.connect.html">
pyarrow.hdfs.connect
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.cat.html">
pyarrow.HadoopFileSystem.cat
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.chmod.html">
pyarrow.HadoopFileSystem.chmod
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.chown.html">
pyarrow.HadoopFileSystem.chown
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.delete.html">
pyarrow.HadoopFileSystem.delete
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.df.html">
pyarrow.HadoopFileSystem.df
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.disk_usage.html">
pyarrow.HadoopFileSystem.disk_usage
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.download.html">
pyarrow.HadoopFileSystem.download
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.exists.html">
pyarrow.HadoopFileSystem.exists
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.get_capacity.html">
pyarrow.HadoopFileSystem.get_capacity
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.get_space_used.html">
pyarrow.HadoopFileSystem.get_space_used
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.info.html">
pyarrow.HadoopFileSystem.info
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.ls.html">
pyarrow.HadoopFileSystem.ls
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.mkdir.html">
pyarrow.HadoopFileSystem.mkdir
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.open.html">
pyarrow.HadoopFileSystem.open
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.rename.html">
pyarrow.HadoopFileSystem.rename
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.rm.html">
pyarrow.HadoopFileSystem.rm
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HadoopFileSystem.upload.html">
pyarrow.HadoopFileSystem.upload
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/generated/pyarrow.HdfsFile.html">
pyarrow.HdfsFile
</a>
</li>
</ul>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/plasma.html">
The Plasma In-Memory Object Store
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/numpy.html">
NumPy Integration
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/pandas.html">
Pandas Integration
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/timestamps.html">
Timestamps
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/orc.html">
Reading and Writing the Apache ORC Format
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/csv.html">
Reading and Writing CSV files
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/feather.html">
Feather File Format
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/json.html">
Reading JSON files
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/parquet.html">
Reading and Writing the Apache Parquet Format
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/dataset.html">
Tabular Datasets
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/flight.html">
Arrow Flight RPC
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/extending_types.html">
Extending pyarrow
</a>
</li>
<li class="toctree-l2 has-children">
<a class="reference internal" href="../python/integration.html">
PyArrow Integrations
</a>
<input class="toctree-checkbox" id="toctree-checkbox-8" name="toctree-checkbox-8" type="checkbox"/>
<label for="toctree-checkbox-8">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l3">
<a class="reference internal" href="../python/integration/python_r.html">
Integrating PyArrow with R
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/integration/python_java.html">
Integrating PyArrow with Java
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/integration/extending.html">
Using pyarrow from C++ and Cython Code
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../python/integration/cuda.html">
CUDA Integration
</a>
</li>
</ul>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/env_vars.html">
Environment Variables
</a>
</li>
<li class="toctree-l2 has-children">
<a class="reference internal" href="../python/api.html">
API Reference
</a>
<input class="toctree-checkbox" id="toctree-checkbox-9" name="toctree-checkbox-9" type="checkbox"/>
<label for="toctree-checkbox-9">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l3 has-children">
<a class="reference internal" href="../python/api/datatypes.html">
Data Types and Schemas
</a>
<input class="toctree-checkbox" id="toctree-checkbox-10" name="toctree-checkbox-10" type="checkbox"/>
<label for="toctree-checkbox-10">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.null.html">
pyarrow.null
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.bool_.html">
pyarrow.bool_
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.int8.html">
pyarrow.int8
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.int16.html">
pyarrow.int16
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.int32.html">
pyarrow.int32
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.int64.html">
pyarrow.int64
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.uint8.html">
pyarrow.uint8
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.uint16.html">
pyarrow.uint16
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.uint32.html">
pyarrow.uint32
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.uint64.html">
pyarrow.uint64
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.float16.html">
pyarrow.float16
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.float32.html">
pyarrow.float32
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.float64.html">
pyarrow.float64
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.time32.html">
pyarrow.time32
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.time64.html">
pyarrow.time64
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.timestamp.html">
pyarrow.timestamp
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.date32.html">
pyarrow.date32
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.date64.html">
pyarrow.date64
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.duration.html">
pyarrow.duration
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.month_day_nano_interval.html">
pyarrow.month_day_nano_interval
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.binary.html">
pyarrow.binary
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.string.html">
pyarrow.string
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.utf8.html">
pyarrow.utf8
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.large_binary.html">
pyarrow.large_binary
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.large_string.html">
pyarrow.large_string
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.large_utf8.html">
pyarrow.large_utf8
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.decimal128.html">
pyarrow.decimal128
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.list_.html">
pyarrow.list_
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.large_list.html">
pyarrow.large_list
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.map_.html">
pyarrow.map_
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.struct.html">
pyarrow.struct
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dictionary.html">
pyarrow.dictionary
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.field.html">
pyarrow.field
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.schema.html">
pyarrow.schema
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.from_numpy_dtype.html">
pyarrow.from_numpy_dtype
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.unify_schemas.html">
pyarrow.unify_schemas
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.DataType.html">
pyarrow.DataType
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.DictionaryType.html">
pyarrow.DictionaryType
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ListType.html">
pyarrow.ListType
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.MapType.html">
pyarrow.MapType
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.StructType.html">
pyarrow.StructType
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.UnionType.html">
pyarrow.UnionType
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.TimestampType.html">
pyarrow.TimestampType
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Time32Type.html">
pyarrow.Time32Type
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Time64Type.html">
pyarrow.Time64Type
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.FixedSizeBinaryType.html">
pyarrow.FixedSizeBinaryType
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Decimal128Type.html">
pyarrow.Decimal128Type
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Field.html">
pyarrow.Field
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Schema.html">
pyarrow.Schema
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ExtensionType.html">
pyarrow.ExtensionType
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.PyExtensionType.html">
pyarrow.PyExtensionType
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.register_extension_type.html">
pyarrow.register_extension_type
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.unregister_extension_type.html">
pyarrow.unregister_extension_type
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_boolean.html">
pyarrow.types.is_boolean
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_integer.html">
pyarrow.types.is_integer
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_signed_integer.html">
pyarrow.types.is_signed_integer
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_unsigned_integer.html">
pyarrow.types.is_unsigned_integer
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_int8.html">
pyarrow.types.is_int8
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_int16.html">
pyarrow.types.is_int16
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_int32.html">
pyarrow.types.is_int32
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_int64.html">
pyarrow.types.is_int64
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_uint8.html">
pyarrow.types.is_uint8
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_uint16.html">
pyarrow.types.is_uint16
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_uint32.html">
pyarrow.types.is_uint32
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_uint64.html">
pyarrow.types.is_uint64
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_floating.html">
pyarrow.types.is_floating
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_float16.html">
pyarrow.types.is_float16
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_float32.html">
pyarrow.types.is_float32
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_float64.html">
pyarrow.types.is_float64
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_decimal.html">
pyarrow.types.is_decimal
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_list.html">
pyarrow.types.is_list
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_large_list.html">
pyarrow.types.is_large_list
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_struct.html">
pyarrow.types.is_struct
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_union.html">
pyarrow.types.is_union
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_nested.html">
pyarrow.types.is_nested
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_temporal.html">
pyarrow.types.is_temporal
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_timestamp.html">
pyarrow.types.is_timestamp
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_date.html">
pyarrow.types.is_date
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_date32.html">
pyarrow.types.is_date32
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_date64.html">
pyarrow.types.is_date64
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_time.html">
pyarrow.types.is_time
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_time32.html">
pyarrow.types.is_time32
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_time64.html">
pyarrow.types.is_time64
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_null.html">
pyarrow.types.is_null
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_binary.html">
pyarrow.types.is_binary
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_unicode.html">
pyarrow.types.is_unicode
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_string.html">
pyarrow.types.is_string
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_large_binary.html">
pyarrow.types.is_large_binary
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_large_unicode.html">
pyarrow.types.is_large_unicode
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_large_string.html">
pyarrow.types.is_large_string
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_fixed_size_binary.html">
pyarrow.types.is_fixed_size_binary
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_map.html">
pyarrow.types.is_map
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.types.is_dictionary.html">
pyarrow.types.is_dictionary
</a>
</li>
</ul>
</li>
<li class="toctree-l3 has-children">
<a class="reference internal" href="../python/api/arrays.html">
Arrays and Scalars
</a>
<input class="toctree-checkbox" id="toctree-checkbox-11" name="toctree-checkbox-11" type="checkbox"/>
<label for="toctree-checkbox-11">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.array.html">
pyarrow.array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.nulls.html">
pyarrow.nulls
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Array.html">
pyarrow.Array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.BooleanArray.html">
pyarrow.BooleanArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.FloatingPointArray.html">
pyarrow.FloatingPointArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.IntegerArray.html">
pyarrow.IntegerArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Int8Array.html">
pyarrow.Int8Array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Int16Array.html">
pyarrow.Int16Array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Int32Array.html">
pyarrow.Int32Array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Int64Array.html">
pyarrow.Int64Array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.NullArray.html">
pyarrow.NullArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.NumericArray.html">
pyarrow.NumericArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.UInt8Array.html">
pyarrow.UInt8Array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.UInt16Array.html">
pyarrow.UInt16Array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.UInt32Array.html">
pyarrow.UInt32Array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.UInt64Array.html">
pyarrow.UInt64Array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.BinaryArray.html">
pyarrow.BinaryArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.StringArray.html">
pyarrow.StringArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.FixedSizeBinaryArray.html">
pyarrow.FixedSizeBinaryArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.LargeBinaryArray.html">
pyarrow.LargeBinaryArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.LargeStringArray.html">
pyarrow.LargeStringArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Time32Array.html">
pyarrow.Time32Array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Time64Array.html">
pyarrow.Time64Array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Date32Array.html">
pyarrow.Date32Array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Date64Array.html">
pyarrow.Date64Array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.TimestampArray.html">
pyarrow.TimestampArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.DurationArray.html">
pyarrow.DurationArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.MonthDayNanoIntervalArray.html">
pyarrow.MonthDayNanoIntervalArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Decimal128Array.html">
pyarrow.Decimal128Array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.DictionaryArray.html">
pyarrow.DictionaryArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ListArray.html">
pyarrow.ListArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.FixedSizeListArray.html">
pyarrow.FixedSizeListArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.LargeListArray.html">
pyarrow.LargeListArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.MapArray.html">
pyarrow.MapArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.StructArray.html">
pyarrow.StructArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.UnionArray.html">
pyarrow.UnionArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ExtensionArray.html">
pyarrow.ExtensionArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.scalar.html">
pyarrow.scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.NA.html">
pyarrow.NA
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Scalar.html">
pyarrow.Scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.BooleanScalar.html">
pyarrow.BooleanScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Int8Scalar.html">
pyarrow.Int8Scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Int16Scalar.html">
pyarrow.Int16Scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Int32Scalar.html">
pyarrow.Int32Scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Int64Scalar.html">
pyarrow.Int64Scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.UInt8Scalar.html">
pyarrow.UInt8Scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.UInt16Scalar.html">
pyarrow.UInt16Scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.UInt32Scalar.html">
pyarrow.UInt32Scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.UInt64Scalar.html">
pyarrow.UInt64Scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.FloatScalar.html">
pyarrow.FloatScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.DoubleScalar.html">
pyarrow.DoubleScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.BinaryScalar.html">
pyarrow.BinaryScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.StringScalar.html">
pyarrow.StringScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.FixedSizeBinaryScalar.html">
pyarrow.FixedSizeBinaryScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.LargeBinaryScalar.html">
pyarrow.LargeBinaryScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.LargeStringScalar.html">
pyarrow.LargeStringScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Time32Scalar.html">
pyarrow.Time32Scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Time64Scalar.html">
pyarrow.Time64Scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Date32Scalar.html">
pyarrow.Date32Scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Date64Scalar.html">
pyarrow.Date64Scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.TimestampScalar.html">
pyarrow.TimestampScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.DurationScalar.html">
pyarrow.DurationScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.MonthDayNanoIntervalScalar.html">
pyarrow.MonthDayNanoIntervalScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Decimal128Scalar.html">
pyarrow.Decimal128Scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.DictionaryScalar.html">
pyarrow.DictionaryScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ListScalar.html">
pyarrow.ListScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.LargeListScalar.html">
pyarrow.LargeListScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.MapScalar.html">
pyarrow.MapScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.StructScalar.html">
pyarrow.StructScalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.UnionScalar.html">
pyarrow.UnionScalar
</a>
</li>
</ul>
</li>
<li class="toctree-l3 has-children">
<a class="reference internal" href="../python/api/memory.html">
Buffers and Memory
</a>
<input class="toctree-checkbox" id="toctree-checkbox-12" name="toctree-checkbox-12" type="checkbox"/>
<label for="toctree-checkbox-12">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.allocate_buffer.html">
pyarrow.allocate_buffer
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.py_buffer.html">
pyarrow.py_buffer
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.foreign_buffer.html">
pyarrow.foreign_buffer
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Buffer.html">
pyarrow.Buffer
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ResizableBuffer.html">
pyarrow.ResizableBuffer
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Codec.html">
pyarrow.Codec
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compress.html">
pyarrow.compress
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.decompress.html">
pyarrow.decompress
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.MemoryPool.html">
pyarrow.MemoryPool
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.default_memory_pool.html">
pyarrow.default_memory_pool
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.jemalloc_memory_pool.html">
pyarrow.jemalloc_memory_pool
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.mimalloc_memory_pool.html">
pyarrow.mimalloc_memory_pool
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.system_memory_pool.html">
pyarrow.system_memory_pool
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.jemalloc_set_decay_ms.html">
pyarrow.jemalloc_set_decay_ms
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.set_memory_pool.html">
pyarrow.set_memory_pool
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.log_memory_allocations.html">
pyarrow.log_memory_allocations
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.total_allocated_bytes.html">
pyarrow.total_allocated_bytes
</a>
</li>
</ul>
</li>
<li class="toctree-l3 has-children">
<a class="reference internal" href="../python/api/compute.html">
Compute Functions
</a>
<input class="toctree-checkbox" id="toctree-checkbox-13" name="toctree-checkbox-13" type="checkbox"/>
<label for="toctree-checkbox-13">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.all.html">
pyarrow.compute.all
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.any.html">
pyarrow.compute.any
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.approximate_median.html">
pyarrow.compute.approximate_median
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.count.html">
pyarrow.compute.count
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.count_distinct.html">
pyarrow.compute.count_distinct
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.index.html">
pyarrow.compute.index
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.max.html">
pyarrow.compute.max
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.mean.html">
pyarrow.compute.mean
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.min.html">
pyarrow.compute.min
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.min_max.html">
pyarrow.compute.min_max
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.mode.html">
pyarrow.compute.mode
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.product.html">
pyarrow.compute.product
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.quantile.html">
pyarrow.compute.quantile
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.stddev.html">
pyarrow.compute.stddev
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.sum.html">
pyarrow.compute.sum
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.tdigest.html">
pyarrow.compute.tdigest
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.variance.html">
pyarrow.compute.variance
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.abs.html">
pyarrow.compute.abs
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.abs_checked.html">
pyarrow.compute.abs_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.add.html">
pyarrow.compute.add
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.add_checked.html">
pyarrow.compute.add_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.divide.html">
pyarrow.compute.divide
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.divide_checked.html">
pyarrow.compute.divide_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.multiply.html">
pyarrow.compute.multiply
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.multiply_checked.html">
pyarrow.compute.multiply_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.negate.html">
pyarrow.compute.negate
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.negate_checked.html">
pyarrow.compute.negate_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.power.html">
pyarrow.compute.power
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.power_checked.html">
pyarrow.compute.power_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.sign.html">
pyarrow.compute.sign
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.sqrt.html">
pyarrow.compute.sqrt
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.sqrt_checked.html">
pyarrow.compute.sqrt_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.subtract.html">
pyarrow.compute.subtract
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.subtract_checked.html">
pyarrow.compute.subtract_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.bit_wise_and.html">
pyarrow.compute.bit_wise_and
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.bit_wise_not.html">
pyarrow.compute.bit_wise_not
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.bit_wise_or.html">
pyarrow.compute.bit_wise_or
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.bit_wise_xor.html">
pyarrow.compute.bit_wise_xor
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.shift_left.html">
pyarrow.compute.shift_left
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.shift_left_checked.html">
pyarrow.compute.shift_left_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.shift_right.html">
pyarrow.compute.shift_right
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.shift_right_checked.html">
pyarrow.compute.shift_right_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ceil.html">
pyarrow.compute.ceil
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.floor.html">
pyarrow.compute.floor
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.round.html">
pyarrow.compute.round
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.round_to_multiple.html">
pyarrow.compute.round_to_multiple
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.trunc.html">
pyarrow.compute.trunc
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ln.html">
pyarrow.compute.ln
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ln_checked.html">
pyarrow.compute.ln_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.log10.html">
pyarrow.compute.log10
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.log10_checked.html">
pyarrow.compute.log10_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.log1p.html">
pyarrow.compute.log1p
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.log1p_checked.html">
pyarrow.compute.log1p_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.log2.html">
pyarrow.compute.log2
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.log2_checked.html">
pyarrow.compute.log2_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.logb.html">
pyarrow.compute.logb
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.logb_checked.html">
pyarrow.compute.logb_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.acos.html">
pyarrow.compute.acos
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.acos_checked.html">
pyarrow.compute.acos_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.asin.html">
pyarrow.compute.asin
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.asin_checked.html">
pyarrow.compute.asin_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.atan.html">
pyarrow.compute.atan
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.atan2.html">
pyarrow.compute.atan2
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.cos.html">
pyarrow.compute.cos
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.cos_checked.html">
pyarrow.compute.cos_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.sin.html">
pyarrow.compute.sin
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.sin_checked.html">
pyarrow.compute.sin_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.tan.html">
pyarrow.compute.tan
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.tan_checked.html">
pyarrow.compute.tan_checked
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.equal.html">
pyarrow.compute.equal
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.greater.html">
pyarrow.compute.greater
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.greater_equal.html">
pyarrow.compute.greater_equal
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.less.html">
pyarrow.compute.less
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.less_equal.html">
pyarrow.compute.less_equal
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.not_equal.html">
pyarrow.compute.not_equal
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.max_element_wise.html">
pyarrow.compute.max_element_wise
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.min_element_wise.html">
pyarrow.compute.min_element_wise
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.and_.html">
pyarrow.compute.and_
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.and_kleene.html">
pyarrow.compute.and_kleene
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.and_not.html">
pyarrow.compute.and_not
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.and_not_kleene.html">
pyarrow.compute.and_not_kleene
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.invert.html">
pyarrow.compute.invert
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.or_.html">
pyarrow.compute.or_
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.or_kleene.html">
pyarrow.compute.or_kleene
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.xor.html">
pyarrow.compute.xor
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_is_alnum.html">
pyarrow.compute.ascii_is_alnum
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_is_alpha.html">
pyarrow.compute.ascii_is_alpha
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_is_decimal.html">
pyarrow.compute.ascii_is_decimal
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_is_lower.html">
pyarrow.compute.ascii_is_lower
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_is_printable.html">
pyarrow.compute.ascii_is_printable
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_is_space.html">
pyarrow.compute.ascii_is_space
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_is_upper.html">
pyarrow.compute.ascii_is_upper
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_is_alnum.html">
pyarrow.compute.utf8_is_alnum
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_is_alpha.html">
pyarrow.compute.utf8_is_alpha
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_is_decimal.html">
pyarrow.compute.utf8_is_decimal
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_is_digit.html">
pyarrow.compute.utf8_is_digit
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_is_lower.html">
pyarrow.compute.utf8_is_lower
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_is_numeric.html">
pyarrow.compute.utf8_is_numeric
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_is_printable.html">
pyarrow.compute.utf8_is_printable
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_is_space.html">
pyarrow.compute.utf8_is_space
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_is_upper.html">
pyarrow.compute.utf8_is_upper
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_is_title.html">
pyarrow.compute.ascii_is_title
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_is_title.html">
pyarrow.compute.utf8_is_title
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.string_is_ascii.html">
pyarrow.compute.string_is_ascii
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_capitalize.html">
pyarrow.compute.ascii_capitalize
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_lower.html">
pyarrow.compute.ascii_lower
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_reverse.html">
pyarrow.compute.ascii_reverse
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_swapcase.html">
pyarrow.compute.ascii_swapcase
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_title.html">
pyarrow.compute.ascii_title
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_upper.html">
pyarrow.compute.ascii_upper
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.binary_length.html">
pyarrow.compute.binary_length
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.binary_repeat.html">
pyarrow.compute.binary_repeat
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.binary_replace_slice.html">
pyarrow.compute.binary_replace_slice
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.binary_reverse.html">
pyarrow.compute.binary_reverse
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.replace_substring.html">
pyarrow.compute.replace_substring
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.replace_substring_regex.html">
pyarrow.compute.replace_substring_regex
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_capitalize.html">
pyarrow.compute.utf8_capitalize
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_length.html">
pyarrow.compute.utf8_length
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_lower.html">
pyarrow.compute.utf8_lower
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_replace_slice.html">
pyarrow.compute.utf8_replace_slice
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_reverse.html">
pyarrow.compute.utf8_reverse
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_swapcase.html">
pyarrow.compute.utf8_swapcase
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_title.html">
pyarrow.compute.utf8_title
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_upper.html">
pyarrow.compute.utf8_upper
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_center.html">
pyarrow.compute.ascii_center
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_lpad.html">
pyarrow.compute.ascii_lpad
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_rpad.html">
pyarrow.compute.ascii_rpad
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_center.html">
pyarrow.compute.utf8_center
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_lpad.html">
pyarrow.compute.utf8_lpad
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_rpad.html">
pyarrow.compute.utf8_rpad
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_ltrim.html">
pyarrow.compute.ascii_ltrim
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_ltrim_whitespace.html">
pyarrow.compute.ascii_ltrim_whitespace
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_rtrim.html">
pyarrow.compute.ascii_rtrim
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_rtrim_whitespace.html">
pyarrow.compute.ascii_rtrim_whitespace
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_trim.html">
pyarrow.compute.ascii_trim
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_trim_whitespace.html">
pyarrow.compute.ascii_trim_whitespace
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_ltrim.html">
pyarrow.compute.utf8_ltrim
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_ltrim_whitespace.html">
pyarrow.compute.utf8_ltrim_whitespace
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_rtrim.html">
pyarrow.compute.utf8_rtrim
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_rtrim_whitespace.html">
pyarrow.compute.utf8_rtrim_whitespace
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_trim.html">
pyarrow.compute.utf8_trim
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_trim_whitespace.html">
pyarrow.compute.utf8_trim_whitespace
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ascii_split_whitespace.html">
pyarrow.compute.ascii_split_whitespace
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.split_pattern.html">
pyarrow.compute.split_pattern
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.split_pattern_regex.html">
pyarrow.compute.split_pattern_regex
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_split_whitespace.html">
pyarrow.compute.utf8_split_whitespace
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.extract_regex.html">
pyarrow.compute.extract_regex
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.binary_join.html">
pyarrow.compute.binary_join
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.binary_join_element_wise.html">
pyarrow.compute.binary_join_element_wise
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.utf8_slice_codeunits.html">
pyarrow.compute.utf8_slice_codeunits
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.count_substring.html">
pyarrow.compute.count_substring
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.count_substring_regex.html">
pyarrow.compute.count_substring_regex
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ends_with.html">
pyarrow.compute.ends_with
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.find_substring.html">
pyarrow.compute.find_substring
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.find_substring_regex.html">
pyarrow.compute.find_substring_regex
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.index_in.html">
pyarrow.compute.index_in
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.is_in.html">
pyarrow.compute.is_in
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.match_like.html">
pyarrow.compute.match_like
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.match_substring.html">
pyarrow.compute.match_substring
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.match_substring_regex.html">
pyarrow.compute.match_substring_regex
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.starts_with.html">
pyarrow.compute.starts_with
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.indices_nonzero.html">
pyarrow.compute.indices_nonzero
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.is_finite.html">
pyarrow.compute.is_finite
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.is_inf.html">
pyarrow.compute.is_inf
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.is_nan.html">
pyarrow.compute.is_nan
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.is_null.html">
pyarrow.compute.is_null
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.is_valid.html">
pyarrow.compute.is_valid
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.true_unless_null.html">
pyarrow.compute.true_unless_null
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.case_when.html">
pyarrow.compute.case_when
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.choose.html">
pyarrow.compute.choose
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.coalesce.html">
pyarrow.compute.coalesce
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.if_else.html">
pyarrow.compute.if_else
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.cast.html">
pyarrow.compute.cast
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ceil_temporal.html">
pyarrow.compute.ceil_temporal
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.floor_temporal.html">
pyarrow.compute.floor_temporal
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.round_temporal.html">
pyarrow.compute.round_temporal
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.strftime.html">
pyarrow.compute.strftime
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.strptime.html">
pyarrow.compute.strptime
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.day.html">
pyarrow.compute.day
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.day_of_week.html">
pyarrow.compute.day_of_week
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.day_of_year.html">
pyarrow.compute.day_of_year
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.hour.html">
pyarrow.compute.hour
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.iso_week.html">
pyarrow.compute.iso_week
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.iso_year.html">
pyarrow.compute.iso_year
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.iso_calendar.html">
pyarrow.compute.iso_calendar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.is_leap_year.html">
pyarrow.compute.is_leap_year
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.microsecond.html">
pyarrow.compute.microsecond
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.millisecond.html">
pyarrow.compute.millisecond
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.minute.html">
pyarrow.compute.minute
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.month.html">
pyarrow.compute.month
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.nanosecond.html">
pyarrow.compute.nanosecond
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.quarter.html">
pyarrow.compute.quarter
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.second.html">
pyarrow.compute.second
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.subsecond.html">
pyarrow.compute.subsecond
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.us_week.html">
pyarrow.compute.us_week
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.us_year.html">
pyarrow.compute.us_year
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.week.html">
pyarrow.compute.week
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.year.html">
pyarrow.compute.year
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.year_month_day.html">
pyarrow.compute.year_month_day
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.day_time_interval_between.html">
pyarrow.compute.day_time_interval_between
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.days_between.html">
pyarrow.compute.days_between
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.hours_between.html">
pyarrow.compute.hours_between
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.microseconds_between.html">
pyarrow.compute.microseconds_between
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.milliseconds_between.html">
pyarrow.compute.milliseconds_between
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.minutes_between.html">
pyarrow.compute.minutes_between
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.month_day_nano_interval_between.html">
pyarrow.compute.month_day_nano_interval_between
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.month_interval_between.html">
pyarrow.compute.month_interval_between
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.nanoseconds_between.html">
pyarrow.compute.nanoseconds_between
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.quarters_between.html">
pyarrow.compute.quarters_between
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.seconds_between.html">
pyarrow.compute.seconds_between
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.weeks_between.html">
pyarrow.compute.weeks_between
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.years_between.html">
pyarrow.compute.years_between
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.assume_timezone.html">
pyarrow.compute.assume_timezone
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.dictionary_encode.html">
pyarrow.compute.dictionary_encode
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.unique.html">
pyarrow.compute.unique
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.value_counts.html">
pyarrow.compute.value_counts
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.array_filter.html">
pyarrow.compute.array_filter
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.array_take.html">
pyarrow.compute.array_take
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.drop_null.html">
pyarrow.compute.drop_null
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.filter.html">
pyarrow.compute.filter
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.take.html">
pyarrow.compute.take
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.array_sort_indices.html">
pyarrow.compute.array_sort_indices
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.partition_nth_indices.html">
pyarrow.compute.partition_nth_indices
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.select_k_unstable.html">
pyarrow.compute.select_k_unstable
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.sort_indices.html">
pyarrow.compute.sort_indices
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.fill_null_backward.html">
pyarrow.compute.fill_null_backward
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.fill_null_forward.html">
pyarrow.compute.fill_null_forward
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.list_element.html">
pyarrow.compute.list_element
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.list_flatten.html">
pyarrow.compute.list_flatten
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.list_parent_indices.html">
pyarrow.compute.list_parent_indices
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.list_value_length.html">
pyarrow.compute.list_value_length
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.make_struct.html">
pyarrow.compute.make_struct
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.map_lookup.html">
pyarrow.compute.map_lookup
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.replace_with_mask.html">
pyarrow.compute.replace_with_mask
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.struct_field.html">
pyarrow.compute.struct_field
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ArraySortOptions.html">
pyarrow.compute.ArraySortOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.AssumeTimezoneOptions.html">
pyarrow.compute.AssumeTimezoneOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.CastOptions.html">
pyarrow.compute.CastOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.CountOptions.html">
pyarrow.compute.CountOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.CountOptions.html">
pyarrow.compute.CountOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.DayOfWeekOptions.html">
pyarrow.compute.DayOfWeekOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.DictionaryEncodeOptions.html">
pyarrow.compute.DictionaryEncodeOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ElementWiseAggregateOptions.html">
pyarrow.compute.ElementWiseAggregateOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ExtractRegexOptions.html">
pyarrow.compute.ExtractRegexOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.FilterOptions.html">
pyarrow.compute.FilterOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.IndexOptions.html">
pyarrow.compute.IndexOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.JoinOptions.html">
pyarrow.compute.JoinOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.MakeStructOptions.html">
pyarrow.compute.MakeStructOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.MapLookupOptions.html">
pyarrow.compute.MapLookupOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.MatchSubstringOptions.html">
pyarrow.compute.MatchSubstringOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ModeOptions.html">
pyarrow.compute.ModeOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.NullOptions.html">
pyarrow.compute.NullOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.PadOptions.html">
pyarrow.compute.PadOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.PartitionNthOptions.html">
pyarrow.compute.PartitionNthOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.QuantileOptions.html">
pyarrow.compute.QuantileOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ReplaceSliceOptions.html">
pyarrow.compute.ReplaceSliceOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ReplaceSubstringOptions.html">
pyarrow.compute.ReplaceSubstringOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.RoundOptions.html">
pyarrow.compute.RoundOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.RoundTemporalOptions.html">
pyarrow.compute.RoundTemporalOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.RoundToMultipleOptions.html">
pyarrow.compute.RoundToMultipleOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ScalarAggregateOptions.html">
pyarrow.compute.ScalarAggregateOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.ScalarAggregateOptions.html">
pyarrow.compute.ScalarAggregateOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.SelectKOptions.html">
pyarrow.compute.SelectKOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.SetLookupOptions.html">
pyarrow.compute.SetLookupOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.SliceOptions.html">
pyarrow.compute.SliceOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.SortOptions.html">
pyarrow.compute.SortOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.SplitOptions.html">
pyarrow.compute.SplitOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.SplitPatternOptions.html">
pyarrow.compute.SplitPatternOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.StrftimeOptions.html">
pyarrow.compute.StrftimeOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.StrptimeOptions.html">
pyarrow.compute.StrptimeOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.StructFieldOptions.html">
pyarrow.compute.StructFieldOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.TakeOptions.html">
pyarrow.compute.TakeOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.TDigestOptions.html">
pyarrow.compute.TDigestOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.TDigestOptions.html">
pyarrow.compute.TDigestOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.TrimOptions.html">
pyarrow.compute.TrimOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.VarianceOptions.html">
pyarrow.compute.VarianceOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.compute.WeekOptions.html">
pyarrow.compute.WeekOptions
</a>
</li>
</ul>
</li>
<li class="toctree-l3 has-children">
<a class="reference internal" href="../python/api/files.html">
Streams and File Access
</a>
<input class="toctree-checkbox" id="toctree-checkbox-14" name="toctree-checkbox-14" type="checkbox"/>
<label for="toctree-checkbox-14">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.input_stream.html">
pyarrow.input_stream
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.output_stream.html">
pyarrow.output_stream
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.memory_map.html">
pyarrow.memory_map
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.create_memory_map.html">
pyarrow.create_memory_map
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.NativeFile.html">
pyarrow.NativeFile
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.OSFile.html">
pyarrow.OSFile
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.PythonFile.html">
pyarrow.PythonFile
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.BufferReader.html">
pyarrow.BufferReader
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.BufferOutputStream.html">
pyarrow.BufferOutputStream
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.FixedSizeBufferWriter.html">
pyarrow.FixedSizeBufferWriter
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.MemoryMappedFile.html">
pyarrow.MemoryMappedFile
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.CompressedInputStream.html">
pyarrow.CompressedInputStream
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.CompressedOutputStream.html">
pyarrow.CompressedOutputStream
</a>
</li>
</ul>
</li>
<li class="toctree-l3 has-children">
<a class="reference internal" href="../python/api/tables.html">
Tables and Tensors
</a>
<input class="toctree-checkbox" id="toctree-checkbox-15" name="toctree-checkbox-15" type="checkbox"/>
<label for="toctree-checkbox-15">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.chunked_array.html">
pyarrow.chunked_array
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.concat_arrays.html">
pyarrow.concat_arrays
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.concat_tables.html">
pyarrow.concat_tables
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.record_batch.html">
pyarrow.record_batch
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.table.html">
pyarrow.table
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ChunkedArray.html">
pyarrow.ChunkedArray
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.RecordBatch.html">
pyarrow.RecordBatch
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Table.html">
pyarrow.Table
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.TableGroupBy.html">
pyarrow.TableGroupBy
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.RecordBatchReader.html">
pyarrow.RecordBatchReader
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.Tensor.html">
pyarrow.Tensor
</a>
</li>
</ul>
</li>
<li class="toctree-l3 has-children">
<a class="reference internal" href="../python/api/ipc.html">
Serialization and IPC
</a>
<input class="toctree-checkbox" id="toctree-checkbox-16" name="toctree-checkbox-16" type="checkbox"/>
<label for="toctree-checkbox-16">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.new_file.html">
pyarrow.ipc.new_file
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.open_file.html">
pyarrow.ipc.open_file
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.new_stream.html">
pyarrow.ipc.new_stream
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.open_stream.html">
pyarrow.ipc.open_stream
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.read_message.html">
pyarrow.ipc.read_message
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.read_record_batch.html">
pyarrow.ipc.read_record_batch
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.get_record_batch_size.html">
pyarrow.ipc.get_record_batch_size
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.read_tensor.html">
pyarrow.ipc.read_tensor
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.write_tensor.html">
pyarrow.ipc.write_tensor
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.get_tensor_size.html">
pyarrow.ipc.get_tensor_size
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.IpcReadOptions.html">
pyarrow.ipc.IpcReadOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.IpcWriteOptions.html">
pyarrow.ipc.IpcWriteOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.Message.html">
pyarrow.ipc.Message
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.MessageReader.html">
pyarrow.ipc.MessageReader
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.RecordBatchFileReader.html">
pyarrow.ipc.RecordBatchFileReader
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.RecordBatchFileWriter.html">
pyarrow.ipc.RecordBatchFileWriter
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.RecordBatchStreamReader.html">
pyarrow.ipc.RecordBatchStreamReader
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.ipc.RecordBatchStreamWriter.html">
pyarrow.ipc.RecordBatchStreamWriter
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.serialize.html">
pyarrow.serialize
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.serialize_to.html">
pyarrow.serialize_to
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.deserialize.html">
pyarrow.deserialize
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.deserialize_components.html">
pyarrow.deserialize_components
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.deserialize_from.html">
pyarrow.deserialize_from
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.read_serialized.html">
pyarrow.read_serialized
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.SerializedPyObject.html">
pyarrow.SerializedPyObject
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.SerializationContext.html">
pyarrow.SerializationContext
</a>
</li>
</ul>
</li>
<li class="toctree-l3 has-children">
<a class="reference internal" href="../python/api/flight.html">
Arrow Flight
</a>
<input class="toctree-checkbox" id="toctree-checkbox-17" name="toctree-checkbox-17" type="checkbox"/>
<label for="toctree-checkbox-17">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.Action.html">
pyarrow.flight.Action
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.ActionType.html">
pyarrow.flight.ActionType
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.DescriptorType.html">
pyarrow.flight.DescriptorType
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightDescriptor.html">
pyarrow.flight.FlightDescriptor
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightEndpoint.html">
pyarrow.flight.FlightEndpoint
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightInfo.html">
pyarrow.flight.FlightInfo
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.Location.html">
pyarrow.flight.Location
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.MetadataRecordBatchReader.html">
pyarrow.flight.MetadataRecordBatchReader
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.MetadataRecordBatchWriter.html">
pyarrow.flight.MetadataRecordBatchWriter
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.Ticket.html">
pyarrow.flight.Ticket
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.Result.html">
pyarrow.flight.Result
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.connect.html">
pyarrow.flight.connect
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightCallOptions.html">
pyarrow.flight.FlightCallOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightClient.html">
pyarrow.flight.FlightClient
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightStreamReader.html">
pyarrow.flight.FlightStreamReader
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightStreamWriter.html">
pyarrow.flight.FlightStreamWriter
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.ClientMiddlewareFactory.html">
pyarrow.flight.ClientMiddlewareFactory
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.ClientMiddleware.html">
pyarrow.flight.ClientMiddleware
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightDataStream.html">
pyarrow.flight.FlightDataStream
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightMetadataWriter.html">
pyarrow.flight.FlightMetadataWriter
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightServerBase.html">
pyarrow.flight.FlightServerBase
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.GeneratorStream.html">
pyarrow.flight.GeneratorStream
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.RecordBatchStream.html">
pyarrow.flight.RecordBatchStream
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.ServerCallContext.html">
pyarrow.flight.ServerCallContext
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.ServerMiddlewareFactory.html">
pyarrow.flight.ServerMiddlewareFactory
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.ServerMiddleware.html">
pyarrow.flight.ServerMiddleware
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.ClientAuthHandler.html">
pyarrow.flight.ClientAuthHandler
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.ServerAuthHandler.html">
pyarrow.flight.ServerAuthHandler
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightError.html">
pyarrow.flight.FlightError
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightCancelledError.html">
pyarrow.flight.FlightCancelledError
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightInternalError.html">
pyarrow.flight.FlightInternalError
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightServerError.html">
pyarrow.flight.FlightServerError
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightTimedOutError.html">
pyarrow.flight.FlightTimedOutError
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightUnauthenticatedError.html">
pyarrow.flight.FlightUnauthenticatedError
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightUnauthorizedError.html">
pyarrow.flight.FlightUnauthorizedError
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightUnavailableError.html">
pyarrow.flight.FlightUnavailableError
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightWriteSizeExceededError.html">
pyarrow.flight.FlightWriteSizeExceededError
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.FlightMethod.html">
pyarrow.flight.FlightMethod
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.flight.CallInfo.html">
pyarrow.flight.CallInfo
</a>
</li>
</ul>
</li>
<li class="toctree-l3 has-children">
<a class="reference internal" href="../python/api/formats.html">
Tabular File Formats
</a>
<input class="toctree-checkbox" id="toctree-checkbox-18" name="toctree-checkbox-18" type="checkbox"/>
<label for="toctree-checkbox-18">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.csv.ConvertOptions.html">
pyarrow.csv.ConvertOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.csv.CSVStreamingReader.html">
pyarrow.csv.CSVStreamingReader
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.csv.CSVWriter.html">
pyarrow.csv.CSVWriter
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.csv.ISO8601.html">
pyarrow.csv.ISO8601
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.csv.ParseOptions.html">
pyarrow.csv.ParseOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.csv.ReadOptions.html">
pyarrow.csv.ReadOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.csv.WriteOptions.html">
pyarrow.csv.WriteOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.csv.open_csv.html">
pyarrow.csv.open_csv
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.csv.read_csv.html">
pyarrow.csv.read_csv
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.csv.write_csv.html">
pyarrow.csv.write_csv
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.csv.InvalidRow.html">
pyarrow.csv.InvalidRow
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.feather.read_feather.html">
pyarrow.feather.read_feather
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.feather.read_table.html">
pyarrow.feather.read_table
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.feather.write_feather.html">
pyarrow.feather.write_feather
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.json.ReadOptions.html">
pyarrow.json.ReadOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.json.ParseOptions.html">
pyarrow.json.ParseOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.json.read_json.html">
pyarrow.json.read_json
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.ParquetDataset.html">
pyarrow.parquet.ParquetDataset
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.ParquetFile.html">
pyarrow.parquet.ParquetFile
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.ParquetWriter.html">
pyarrow.parquet.ParquetWriter
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.read_table.html">
pyarrow.parquet.read_table
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.read_metadata.html">
pyarrow.parquet.read_metadata
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.read_pandas.html">
pyarrow.parquet.read_pandas
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.read_schema.html">
pyarrow.parquet.read_schema
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.write_metadata.html">
pyarrow.parquet.write_metadata
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.write_table.html">
pyarrow.parquet.write_table
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.write_to_dataset.html">
pyarrow.parquet.write_to_dataset
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.FileMetaData.html">
pyarrow.parquet.FileMetaData
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.RowGroupMetaData.html">
pyarrow.parquet.RowGroupMetaData
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.ColumnChunkMetaData.html">
pyarrow.parquet.ColumnChunkMetaData
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.Statistics.html">
pyarrow.parquet.Statistics
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.ParquetSchema.html">
pyarrow.parquet.ParquetSchema
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.ColumnSchema.html">
pyarrow.parquet.ColumnSchema
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.ParquetLogicalType.html">
pyarrow.parquet.ParquetLogicalType
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.encryption.CryptoFactory.html">
pyarrow.parquet.encryption.CryptoFactory
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.encryption.KmsClient.html">
pyarrow.parquet.encryption.KmsClient
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.encryption.KmsConnectionConfig.html">
pyarrow.parquet.encryption.KmsConnectionConfig
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.encryption.EncryptionConfiguration.html">
pyarrow.parquet.encryption.EncryptionConfiguration
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.parquet.encryption.DecryptionConfiguration.html">
pyarrow.parquet.encryption.DecryptionConfiguration
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.orc.ORCFile.html">
pyarrow.orc.ORCFile
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.orc.ORCWriter.html">
pyarrow.orc.ORCWriter
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.orc.read_table.html">
pyarrow.orc.read_table
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.orc.write_table.html">
pyarrow.orc.write_table
</a>
</li>
</ul>
</li>
<li class="toctree-l3 has-children">
<a class="reference internal" href="../python/api/filesystems.html">
Filesystems
</a>
<input class="toctree-checkbox" id="toctree-checkbox-19" name="toctree-checkbox-19" type="checkbox"/>
<label for="toctree-checkbox-19">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.FileInfo.html">
pyarrow.fs.FileInfo
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.FileSelector.html">
pyarrow.fs.FileSelector
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.FileSystem.html">
pyarrow.fs.FileSystem
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.LocalFileSystem.html">
pyarrow.fs.LocalFileSystem
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.S3FileSystem.html">
pyarrow.fs.S3FileSystem
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.HadoopFileSystem.html">
pyarrow.fs.HadoopFileSystem
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.SubTreeFileSystem.html">
pyarrow.fs.SubTreeFileSystem
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.PyFileSystem.html">
pyarrow.fs.PyFileSystem
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.FileSystemHandler.html">
pyarrow.fs.FileSystemHandler
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.FSSpecHandler.html">
pyarrow.fs.FSSpecHandler
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.copy_files.html">
pyarrow.fs.copy_files
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.initialize_s3.html">
pyarrow.fs.initialize_s3
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.finalize_s3.html">
pyarrow.fs.finalize_s3
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.resolve_s3_region.html">
pyarrow.fs.resolve_s3_region
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.fs.S3LogLevel.html">
pyarrow.fs.S3LogLevel
</a>
</li>
</ul>
</li>
<li class="toctree-l3 has-children">
<a class="reference internal" href="../python/api/dataset.html">
Dataset
</a>
<input class="toctree-checkbox" id="toctree-checkbox-20" name="toctree-checkbox-20" type="checkbox"/>
<label for="toctree-checkbox-20">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.dataset.html">
pyarrow.dataset.dataset
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.parquet_dataset.html">
pyarrow.dataset.parquet_dataset
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.partitioning.html">
pyarrow.dataset.partitioning
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.field.html">
pyarrow.dataset.field
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.scalar.html">
pyarrow.dataset.scalar
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.write_dataset.html">
pyarrow.dataset.write_dataset
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.FileFormat.html">
pyarrow.dataset.FileFormat
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.CsvFileFormat.html">
pyarrow.dataset.CsvFileFormat
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.CsvFragmentScanOptions.html">
pyarrow.dataset.CsvFragmentScanOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.IpcFileFormat.html">
pyarrow.dataset.IpcFileFormat
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.ParquetFileFormat.html">
pyarrow.dataset.ParquetFileFormat
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.ParquetReadOptions.html">
pyarrow.dataset.ParquetReadOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.ParquetFragmentScanOptions.html">
pyarrow.dataset.ParquetFragmentScanOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.OrcFileFormat.html">
pyarrow.dataset.OrcFileFormat
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.Partitioning.html">
pyarrow.dataset.Partitioning
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.PartitioningFactory.html">
pyarrow.dataset.PartitioningFactory
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.DirectoryPartitioning.html">
pyarrow.dataset.DirectoryPartitioning
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.HivePartitioning.html">
pyarrow.dataset.HivePartitioning
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.FilenamePartitioning.html">
pyarrow.dataset.FilenamePartitioning
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.Dataset.html">
pyarrow.dataset.Dataset
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.FileSystemDataset.html">
pyarrow.dataset.FileSystemDataset
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.FileSystemFactoryOptions.html">
pyarrow.dataset.FileSystemFactoryOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.FileSystemDatasetFactory.html">
pyarrow.dataset.FileSystemDatasetFactory
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.UnionDataset.html">
pyarrow.dataset.UnionDataset
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.Fragment.html">
pyarrow.dataset.Fragment
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.FragmentScanOptions.html">
pyarrow.dataset.FragmentScanOptions
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.TaggedRecordBatch.html">
pyarrow.dataset.TaggedRecordBatch
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.Scanner.html">
pyarrow.dataset.Scanner
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.Expression.html">
pyarrow.dataset.Expression
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.dataset.InMemoryDataset.html">
pyarrow.dataset.InMemoryDataset
</a>
</li>
</ul>
</li>
<li class="toctree-l3 has-children">
<a class="reference internal" href="../python/api/plasma.html">
Plasma In-Memory Object Store
</a>
<input class="toctree-checkbox" id="toctree-checkbox-21" name="toctree-checkbox-21" type="checkbox"/>
<label for="toctree-checkbox-21">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.plasma.ObjectID.html">
pyarrow.plasma.ObjectID
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.plasma.PlasmaClient.html">
pyarrow.plasma.PlasmaClient
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.plasma.PlasmaBuffer.html">
pyarrow.plasma.PlasmaBuffer
</a>
</li>
</ul>
</li>
<li class="toctree-l3 has-children">
<a class="reference internal" href="../python/api/cuda.html">
CUDA Integration
</a>
<input class="toctree-checkbox" id="toctree-checkbox-22" name="toctree-checkbox-22" type="checkbox"/>
<label for="toctree-checkbox-22">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.cuda.Context.html">
pyarrow.cuda.Context
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.cuda.CudaBuffer.html">
pyarrow.cuda.CudaBuffer
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.cuda.new_host_buffer.html">
pyarrow.cuda.new_host_buffer
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.cuda.HostBuffer.html">
pyarrow.cuda.HostBuffer
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.cuda.BufferReader.html">
pyarrow.cuda.BufferReader
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.cuda.BufferWriter.html">
pyarrow.cuda.BufferWriter
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.cuda.serialize_record_batch.html">
pyarrow.cuda.serialize_record_batch
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.cuda.read_record_batch.html">
pyarrow.cuda.read_record_batch
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.cuda.read_message.html">
pyarrow.cuda.read_message
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.cuda.IpcMemHandle.html">
pyarrow.cuda.IpcMemHandle
</a>
</li>
</ul>
</li>
<li class="toctree-l3 has-children">
<a class="reference internal" href="../python/api/misc.html">
Miscellaneous
</a>
<input class="toctree-checkbox" id="toctree-checkbox-23" name="toctree-checkbox-23" type="checkbox"/>
<label for="toctree-checkbox-23">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.cpu_count.html">
pyarrow.cpu_count
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.set_cpu_count.html">
pyarrow.set_cpu_count
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.io_thread_count.html">
pyarrow.io_thread_count
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.set_io_thread_count.html">
pyarrow.set_io_thread_count
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.get_include.html">
pyarrow.get_include
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.get_libraries.html">
pyarrow.get_libraries
</a>
</li>
<li class="toctree-l4">
<a class="reference internal" href="../python/generated/pyarrow.get_library_dirs.html">
pyarrow.get_library_dirs
</a>
</li>
</ul>
</li>
</ul>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/getting_involved.html">
Getting Involved
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../python/benchmarks.html">
Benchmarks
</a>
</li>
</ul>
</li>
<li class="toctree-l1">
<a class="reference internal" href="../r/index.html">
R
</a>
</li>
<li class="toctree-l1">
<a class="reference external" href="https://github.com/apache/arrow/blob/master/ruby/README.md">
Ruby
</a>
</li>
<li class="toctree-l1">
<a class="reference external" href="https://docs.rs/crate/arrow/">
Rust
</a>
</li>
<li class="toctree-l1">
<a class="reference internal" href="../status.html">
Implementation Status
</a>
</li>
</ul>
<p aria-level="2" class="caption" role="heading">
<span class="caption-text">
Cookbooks
</span>
</p>
<ul class="nav bd-sidenav">
<li class="toctree-l1">
<a class="reference external" href="https://arrow.apache.org/cookbook/cpp/">
C++
</a>
</li>
<li class="toctree-l1">
<a class="reference external" href="https://arrow.apache.org/cookbook/java/">
Java
</a>
</li>
<li class="toctree-l1">
<a class="reference external" href="https://arrow.apache.org/cookbook/py/">
Python
</a>
</li>
<li class="toctree-l1">
<a class="reference external" href="https://arrow.apache.org/cookbook/r/">
R
</a>
</li>
</ul>
<p aria-level="2" class="caption" role="heading">
<span class="caption-text">
Specifications and Protocols
</span>
</p>
<ul class="current nav bd-sidenav">
<li class="toctree-l1">
<a class="reference internal" href="Versioning.html">
Format Versioning and Stability
</a>
</li>
<li class="toctree-l1 current active">
<a class="current reference internal" href="#">
Arrow Columnar Format
</a>
</li>
<li class="toctree-l1">
<a class="reference internal" href="Flight.html">
Arrow Flight RPC
</a>
</li>
<li class="toctree-l1">
<a class="reference internal" href="FlightSql.html">
Arrow Flight SQL
</a>
</li>
<li class="toctree-l1">
<a class="reference internal" href="Integration.html">
Integration Testing
</a>
</li>
<li class="toctree-l1">
<a class="reference internal" href="CDataInterface.html">
The Arrow C data interface
</a>
</li>
<li class="toctree-l1">
<a class="reference internal" href="CStreamInterface.html">
The Arrow C stream interface
</a>
</li>
<li class="toctree-l1">
<a class="reference internal" href="Other.html">
Other Data Structures
</a>
</li>
</ul>
<p aria-level="2" class="caption" role="heading">
<span class="caption-text">
Development
</span>
</p>
<ul class="nav bd-sidenav">
<li class="toctree-l1">
<a class="reference internal" href="../developers/contributing.html">
Contributing to Apache Arrow
</a>
</li>
<li class="toctree-l1">
<a class="reference internal" href="../developers/bug_reports.html">
Report bugs and propose features
</a>
</li>
<li class="toctree-l1 has-children">
<a class="reference internal" href="../developers/guide/index.html">
New Contributor’s Guide
</a>
<input class="toctree-checkbox" id="toctree-checkbox-24" name="toctree-checkbox-24" type="checkbox"/>
<label for="toctree-checkbox-24">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l2">
<a class="reference internal" href="../developers/guide/architectural_overview.html">
Architectural Overview
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../developers/guide/communication.html">
Communication
</a>
</li>
<li class="toctree-l2 has-children">
<a class="reference internal" href="../developers/guide/step_by_step/index.html">
Steps in making your first PR
</a>
<input class="toctree-checkbox" id="toctree-checkbox-25" name="toctree-checkbox-25" type="checkbox"/>
<label for="toctree-checkbox-25">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l3">
<a class="reference internal" href="../developers/guide/step_by_step/set_up.html">
Set up
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../developers/guide/step_by_step/building.html">
Building the Arrow libraries 🏋🏿‍♀️
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../developers/guide/step_by_step/finding_issues.html">
Finding good first issues 🔎
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../developers/guide/step_by_step/arrow_codebase.html">
Working on the Arrow codebase 🧐
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../developers/guide/step_by_step/testing.html">
Testing 🧪
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../developers/guide/step_by_step/styling.html">
Styling 😎
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../developers/guide/step_by_step/pr_lifecycle.html">
Lifecycle of a pull request
</a>
</li>
</ul>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../developers/guide/documentation.html">
Helping with documentation
</a>
</li>
<li class="toctree-l2 has-children">
<a class="reference internal" href="../developers/guide/tutorials/index.html">
Tutorials
</a>
<input class="toctree-checkbox" id="toctree-checkbox-26" name="toctree-checkbox-26" type="checkbox"/>
<label for="toctree-checkbox-26">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l3">
<a class="reference internal" href="../developers/guide/tutorials/python_tutorial.html">
Python tutorial
</a>
</li>
<li class="toctree-l3">
<a class="reference internal" href="../developers/guide/tutorials/r_tutorial.html">
R tutorials
</a>
</li>
</ul>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../developers/guide/resources.html">
Additional information and resources
</a>
</li>
</ul>
</li>
<li class="toctree-l1">
<a class="reference internal" href="../developers/overview.html">
Contributing Overview
</a>
</li>
<li class="toctree-l1">
<a class="reference internal" href="../developers/reviewing.html">
Reviewing contributions
</a>
</li>
<li class="toctree-l1 has-children">
<a class="reference internal" href="../developers/cpp/index.html">
C++ Development
</a>
<input class="toctree-checkbox" id="toctree-checkbox-27" name="toctree-checkbox-27" type="checkbox"/>
<label for="toctree-checkbox-27">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l2">
<a class="reference internal" href="../developers/cpp/building.html">
Building Arrow C++
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../developers/cpp/development.html">
Development Guidelines
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../developers/cpp/windows.html">
Developing on Windows
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../developers/cpp/conventions.html">
Conventions
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../developers/cpp/fuzzing.html">
Fuzzing Arrow C++
</a>
</li>
</ul>
</li>
<li class="toctree-l1 has-children">
<a class="reference internal" href="../developers/java/index.html">
Java Development
</a>
<input class="toctree-checkbox" id="toctree-checkbox-28" name="toctree-checkbox-28" type="checkbox"/>
<label for="toctree-checkbox-28">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l2">
<a class="reference internal" href="../developers/java/building.html">
Building Arrow Java
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../developers/java/development.html">
Development Guidelines
</a>
</li>
</ul>
</li>
<li class="toctree-l1">
<a class="reference internal" href="../developers/python.html">
Python Development
</a>
</li>
<li class="toctree-l1 has-children">
<a class="reference internal" href="../developers/continuous_integration/index.html">
Continuous Integration
</a>
<input class="toctree-checkbox" id="toctree-checkbox-29" name="toctree-checkbox-29" type="checkbox"/>
<label for="toctree-checkbox-29">
<i class="fas fa-chevron-down">
</i>
</label>
<ul>
<li class="toctree-l2">
<a class="reference internal" href="../developers/continuous_integration/overview.html">
Continuous Integration
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../developers/continuous_integration/docker.html">
Running Docker Builds
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../developers/continuous_integration/archery.html">
Daily Development using Archery
</a>
</li>
<li class="toctree-l2">
<a class="reference internal" href="../developers/continuous_integration/crossbow.html">
Packaging and Testing with Crossbow
</a>
</li>
</ul>
</li>
<li class="toctree-l1">
<a class="reference internal" href="../developers/benchmarks.html">
Benchmarks
</a>
</li>
<li class="toctree-l1">
<a class="reference internal" href="../developers/documentation.html">
Building the Documentation
</a>
</li>
<li class="toctree-l1">
<a class="reference internal" href="../developers/computeir.html">
Arrow Compute IR (Intermediate Representation)
</a>
</li>
</ul>
</div>
</nav>
</div>
<div class="sidebar-end-items">
</div>
</div>
<div class="d-none d-xl-block col-xl-2 bd-toc">
<div class="toc-item">
<div class="tocsection onthispage mt-5 pt-1 pb-3">
<i class="fas fa-list"></i> On this page
</div>
<nav id="bd-toc-nav">
<ul class="visible nav section-nav flex-column">
<li class="toc-h2 nav-item toc-entry">
<a class="reference internal nav-link" href="#terminology">
Terminology
</a>
</li>
<li class="toc-h2 nav-item toc-entry">
<a class="reference internal nav-link" href="#physical-memory-layout">
Physical Memory Layout
</a>
<ul class="visible nav section-nav flex-column">
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#buffer-alignment-and-padding">
Buffer Alignment and Padding
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#array-lengths">
Array lengths
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#null-count">
Null count
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#validity-bitmaps">
Validity bitmaps
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#fixed-size-primitive-layout">
Fixed-size Primitive Layout
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#variable-size-binary-layout">
Variable-size Binary Layout
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#variable-size-list-layout">
Variable-size List Layout
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#fixed-size-list-layout">
Fixed-Size List Layout
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#struct-layout">
Struct Layout
</a>
<ul class="nav section-nav flex-column">
<li class="toc-h4 nav-item toc-entry">
<a class="reference internal nav-link" href="#struct-validity">
Struct Validity
</a>
</li>
</ul>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#union-layout">
Union Layout
</a>
<ul class="nav section-nav flex-column">
<li class="toc-h4 nav-item toc-entry">
<a class="reference internal nav-link" href="#dense-union">
Dense Union
</a>
</li>
<li class="toc-h4 nav-item toc-entry">
<a class="reference internal nav-link" href="#sparse-union">
Sparse Union
</a>
</li>
</ul>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#null-layout">
Null Layout
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#dictionary-encoded-layout">
Dictionary-encoded Layout
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#buffer-listing-for-each-layout">
Buffer Listing for Each Layout
</a>
</li>
</ul>
</li>
<li class="toc-h2 nav-item toc-entry">
<a class="reference internal nav-link" href="#logical-types">
Logical Types
</a>
</li>
<li class="toc-h2 nav-item toc-entry">
<a class="reference internal nav-link" href="#serialization-and-interprocess-communication-ipc">
Serialization and Interprocess Communication (IPC)
</a>
<ul class="visible nav section-nav flex-column">
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#encapsulated-message-format">
Encapsulated message format
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#schema-message">
Schema message
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#recordbatch-message">
RecordBatch message
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#byte-order-endianness">
Byte Order (Endianness)
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#ipc-streaming-format">
IPC Streaming Format
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#ipc-file-format">
IPC File Format
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#dictionary-messages">
Dictionary Messages
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#custom-application-metadata">
Custom Application Metadata
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#extension-types">
Extension Types
</a>
</li>
</ul>
</li>
<li class="toc-h2 nav-item toc-entry">
<a class="reference internal nav-link" href="#implementation-guidelines">
Implementation guidelines
</a>
<ul class="visible nav section-nav flex-column">
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#implementing-a-subset-the-spec">
Implementing a subset the spec
</a>
</li>
<li class="toc-h3 nav-item toc-entry">
<a class="reference internal nav-link" href="#extensibility">
Extensibility
</a>
</li>
</ul>
</li>
</ul>
</nav>
</div>
<div class="toc-item">
<div class="tocsection editthispage">
<a href="https://github.com/apache/arrow/edit/master/docs/source/format/Columnar.rst">
<i class="fas fa-pencil-alt"></i> Edit this page
</a>
</div>
</div>
</div>
<main class="col-12 col-md-9 col-xl-7 py-md-5 pl-md-5 pr-md-4 bd-content" role="main">
<div>
<section id="arrow-columnar-format">
<span id="format-columnar"></span><h1>Arrow Columnar Format<a class="headerlink" href="#arrow-columnar-format" title="Permalink to this headline">#</a></h1>
<p><em>Version: 1.0</em></p>
<p>The “Arrow Columnar Format” includes a language-agnostic in-memory
data structure specification, metadata serialization, and a protocol
for serialization and generic data transport.</p>
<p>This document is intended to provide adequate detail to create a new
implementation of the columnar format without the aid of an existing
implementation. We utilize Google’s <a class="reference external" href="http://github.com/google/flatbuffers">Flatbuffers</a> project for
metadata serialization, so it will be necessary to refer to the
project’s <a class="reference external" href="https://github.com/apache/arrow/tree/master/format">Flatbuffers protocol definition files</a>
while reading this document.</p>
<p>The columnar format has some key features:</p>
<ul class="simple">
<li><p>Data adjacency for sequential access (scans)</p></li>
<li><p>O(1) (constant-time) random access</p></li>
<li><p>SIMD and vectorization-friendly</p></li>
<li><p>Relocatable without “pointer swizzling”, allowing for true zero-copy
access in shared memory</p></li>
</ul>
<p>The Arrow columnar format provides analytical performance and data
locality guarantees in exchange for comparatively more expensive
mutation operations. This document is concerned only with in-memory
data representation and serialization details; issues such as
coordinating mutation of data structures are left to be handled by
implementations.</p>
<section id="terminology">
<h2>Terminology<a class="headerlink" href="#terminology" title="Permalink to this headline">#</a></h2>
<p>Since different projects have used different words to describe various
concepts, here is a small glossary to help disambiguate.</p>
<ul class="simple">
<li><p><strong>Array</strong> or <strong>Vector</strong>: a sequence of values with known length all
having the same type. These terms are used interchangeably in
different Arrow implementations, but we use “array” in this
document.</p></li>
<li><p><strong>Slot</strong>: a single logical value in an array of some particular data type</p></li>
<li><p><strong>Buffer</strong> or <strong>Contiguous memory region</strong>: a sequential virtual
address space with a given length. Any byte can be reached via a
single pointer offset less than the region’s length.</p></li>
<li><p><strong>Physical Layout</strong>: The underlying memory layout for an array
without taking into account any value semantics. For example, a
32-bit signed integer array and 32-bit floating point array have the
same layout.</p></li>
<li><p><strong>Parent</strong> and <strong>child arrays</strong>: names to express relationships
between physical value arrays in a nested type structure. For
example, a <code class="docutils literal notranslate"><span class="pre">List&lt;T&gt;</span></code>-type parent array has a T-type array as its
child (see more on lists below).</p></li>
<li><p><strong>Primitive type</strong>: a data type having no child types. This includes
such types as fixed bit-width, variable-size binary, and null types.</p></li>
<li><p><strong>Nested type</strong>: a data type whose full structure depends on one or
more other child types. Two fully-specified nested types are equal
if and only if their child types are equal. For example, <code class="docutils literal notranslate"><span class="pre">List&lt;U&gt;</span></code>
is distinct from <code class="docutils literal notranslate"><span class="pre">List&lt;V&gt;</span></code> iff U and V are different types.</p></li>
<li><p><strong>Logical type</strong>: An application-facing semantic value type that is
implemented using some physical layout. For example, Decimal
values are stored as 16 bytes in a fixed-size binary
layout. Similarly, strings can be stored as <code class="docutils literal notranslate"><span class="pre">List&lt;1-byte&gt;</span></code>. A
timestamp may be stored as 64-bit fixed-size layout.</p></li>
</ul>
</section>
<section id="physical-memory-layout">
<span id="format-layout"></span><h2>Physical Memory Layout<a class="headerlink" href="#physical-memory-layout" title="Permalink to this headline">#</a></h2>
<p>Arrays are defined by a few pieces of metadata and data:</p>
<ul class="simple">
<li><p>A logical data type.</p></li>
<li><p>A sequence of buffers.</p></li>
<li><p>A length as a 64-bit signed integer. Implementations are permitted
to be limited to 32-bit lengths, see more on this below.</p></li>
<li><p>A null count as a 64-bit signed integer.</p></li>
<li><p>An optional <strong>dictionary</strong>, for dictionary-encoded arrays.</p></li>
</ul>
<p>Nested arrays additionally have a sequence of one or more sets of
these items, called the <strong>child arrays</strong>.</p>
<p>Each logical data type has a well-defined physical layout. Here are
the different physical layouts defined by Arrow:</p>
<ul class="simple">
<li><p><strong>Primitive (fixed-size)</strong>: a sequence of values each having the
same byte or bit width</p></li>
<li><p><strong>Variable-size Binary</strong>: a sequence of values each having a variable
byte length. Two variants of this layout are supported using 32-bit
and 64-bit length encoding.</p></li>
<li><p><strong>Fixed-size List</strong>: a nested layout where each value has the same
number of elements taken from a child data type.</p></li>
<li><p><strong>Variable-size List</strong>: a nested layout where each value is a
variable-length sequence of values taken from a child data type. Two
variants of this layout are supported using 32-bit and 64-bit length
encoding.</p></li>
<li><p><strong>Struct</strong>: a nested layout consisting of a collection of named
child <strong>fields</strong> each having the same length but possibly different
types.</p></li>
<li><p><strong>Sparse</strong> and <strong>Dense Union</strong>: a nested layout representing a
sequence of values, each of which can have type chosen from a
collection of child array types.</p></li>
<li><p><strong>Null</strong>: a sequence of all null values, having null logical type</p></li>
</ul>
<p>The Arrow columnar memory layout only applies to <em>data</em> and not
<em>metadata</em>. Implementations are free to represent metadata in-memory
in whichever form is convenient for them. We handle metadata
<strong>serialization</strong> in an implementation-independent way using
<a class="reference external" href="http://github.com/google/flatbuffers">Flatbuffers</a>, detailed below.</p>
<section id="buffer-alignment-and-padding">
<h3>Buffer Alignment and Padding<a class="headerlink" href="#buffer-alignment-and-padding" title="Permalink to this headline">#</a></h3>
<p>Implementations are recommended to allocate memory on aligned
addresses (multiple of 8- or 64-bytes) and pad (overallocate) to a
length that is a multiple of 8 or 64 bytes. When serializing Arrow
data for interprocess communication, these alignment and padding
requirements are enforced. If possible, we suggest that you prefer
using 64-byte alignment and padding. Unless otherwise noted, padded
bytes do not need to have a specific value.</p>
<p>The alignment requirement follows best practices for optimized memory
access:</p>
<ul class="simple">
<li><p>Elements in numeric arrays will be guaranteed to be retrieved via aligned access.</p></li>
<li><p>On some architectures alignment can help limit partially used cache lines.</p></li>
</ul>
<p>The recommendation for 64 byte alignment comes from the <a class="reference external" href="https://software.intel.com/en-us/articles/practical-intel-avx-optimization-on-2nd-generation-intel-core-processors">Intel
performance guide</a> that recommends alignment of memory to match SIMD
register width. The specific padding length was chosen because it
matches the largest SIMD instruction registers available on widely
deployed x86 architecture (Intel AVX-512).</p>
<p>The recommended padding of 64 bytes allows for using <a class="reference external" href="https://software.intel.com/en-us/cpp-compiler-developer-guide-and-reference-introduction-to-the-simd-data-layout-templates">SIMD</a>
instructions consistently in loops without additional conditional
checks. This should allow for simpler, efficient and CPU
cache-friendly code. In other words, we can load the entire 64-byte
buffer into a 512-bit wide SIMD register and get data-level
parallelism on all the columnar values packed into the 64-byte
buffer. Guaranteed padding can also allow certain compilers to
generate more optimized code directly (e.g. One can safely use Intel’s
<code class="docutils literal notranslate"><span class="pre">-qopt-assume-safe-padding</span></code>).</p>
</section>
<section id="array-lengths">
<h3>Array lengths<a class="headerlink" href="#array-lengths" title="Permalink to this headline">#</a></h3>
<p>Array lengths are represented in the Arrow metadata as a 64-bit signed
integer. An implementation of Arrow is considered valid even if it only
supports lengths up to the maximum 32-bit signed integer, though. If using
Arrow in a multi-language environment, we recommend limiting lengths to
2 <sup>31</sup> - 1 elements or less. Larger data sets can be represented using
multiple array chunks.</p>
</section>
<section id="null-count">
<h3>Null count<a class="headerlink" href="#null-count" title="Permalink to this headline">#</a></h3>
<p>The number of null value slots is a property of the physical array and
considered part of the data structure. The null count is represented
in the Arrow metadata as a 64-bit signed integer, as it may be as
large as the array length.</p>
</section>
<section id="validity-bitmaps">
<h3>Validity bitmaps<a class="headerlink" href="#validity-bitmaps" title="Permalink to this headline">#</a></h3>
<p>Any value in an array may be semantically null, whether primitive or nested
type.</p>
<p>All array types, with the exception of union types (more on these later),
utilize a dedicated memory buffer, known as the validity (or “null”) bitmap, to
encode the nullness or non-nullness of each value slot. The validity bitmap
must be large enough to have at least 1 bit for each array slot.</p>
<p>Whether any array slot is valid (non-null) is encoded in the respective bits of
this bitmap. A 1 (set bit) for index <code class="docutils literal notranslate"><span class="pre">j</span></code> indicates that the value is not null,
while a 0 (bit not set) indicates that it is null. Bitmaps are to be
initialized to be all unset at allocation time (this includes padding):</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">is_valid</span><span class="p">[</span><span class="n">j</span><span class="p">]</span> <span class="o">-&gt;</span> <span class="n">bitmap</span><span class="p">[</span><span class="n">j</span> <span class="o">/</span> <span class="mi">8</span><span class="p">]</span> <span class="o">&amp;</span> <span class="p">(</span><span class="mi">1</span> <span class="o">&lt;&lt;</span> <span class="p">(</span><span class="n">j</span> <span class="o">%</span> <span class="mi">8</span><span class="p">))</span>
</pre></div>
</div>
<p>We use <a class="reference external" href="https://en.wikipedia.org/wiki/Bit_numbering">least-significant bit (LSB) numbering</a> (also known as
bit-endianness). This means that within a group of 8 bits, we read
right-to-left:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">values</span> <span class="o">=</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="n">null</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="n">null</span><span class="p">,</span> <span class="mi">3</span><span class="p">]</span>
<span class="n">bitmap</span>
<span class="n">j</span> <span class="n">mod</span> <span class="mi">8</span> <span class="mi">7</span> <span class="mi">6</span> <span class="mi">5</span> <span class="mi">4</span> <span class="mi">3</span> <span class="mi">2</span> <span class="mi">1</span> <span class="mi">0</span>
<span class="mi">0</span> <span class="mi">0</span> <span class="mi">1</span> <span class="mi">0</span> <span class="mi">1</span> <span class="mi">0</span> <span class="mi">1</span> <span class="mi">1</span>
</pre></div>
</div>
<p>Arrays having a 0 null count may choose to not allocate the validity
bitmap; how this is represented depends on the implementation (for
example, a C++ implementation may represent such an “absent” validity
bitmap using a NULL pointer). Implementations may choose to always allocate
a validity bitmap anyway as a matter of convenience. Consumers of Arrow
arrays should be ready to handle those two possibilities.</p>
<p>Nested type arrays (except for union types as noted above) have their own
top-level validity bitmap and null count, regardless of the null count and
valid bits of their child arrays.</p>
<p>Array slots which are null are not required to have a particular value;
any “masked” memory can have any value and need not be zeroed, though
implementations frequently choose to zero memory for null values.</p>
</section>
<section id="fixed-size-primitive-layout">
<h3>Fixed-size Primitive Layout<a class="headerlink" href="#fixed-size-primitive-layout" title="Permalink to this headline">#</a></h3>
<p>A primitive value array represents an array of values each having the
same physical slot width typically measured in bytes, though the spec
also provides for bit-packed types (e.g. boolean values encoded in
bits).</p>
<p>Internally, the array contains a contiguous memory buffer whose total
size is at least as large as the slot width multiplied by the array
length. For bit-packed types, the size is rounded up to the nearest
byte.</p>
<p>The associated validity bitmap is contiguously allocated (as described
above) but does not need to be adjacent in memory to the values
buffer.</p>
<p><strong>Example Layout: Int32 Array</strong></p>
<p>For example a primitive array of int32s:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="p">[</span><span class="mi">1</span><span class="p">,</span> <span class="n">null</span><span class="p">,</span> <span class="mi">2</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">8</span><span class="p">]</span>
</pre></div>
</div>
<p>Would look like:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="o">*</span> <span class="n">Length</span><span class="p">:</span> <span class="mi">5</span><span class="p">,</span> <span class="n">Null</span> <span class="n">count</span><span class="p">:</span> <span class="mi">1</span>
<span class="o">*</span> <span class="n">Validity</span> <span class="n">bitmap</span> <span class="n">buffer</span><span class="p">:</span>
<span class="o">|</span><span class="n">Byte</span> <span class="mi">0</span> <span class="p">(</span><span class="n">validity</span> <span class="n">bitmap</span><span class="p">)</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">1</span><span class="o">-</span><span class="mi">63</span> <span class="o">|</span>
<span class="o">|-------------------------|-----------------------|</span>
<span class="o">|</span> <span class="mi">00011101</span> <span class="o">|</span> <span class="mi">0</span> <span class="p">(</span><span class="n">padding</span><span class="p">)</span> <span class="o">|</span>
<span class="o">*</span> <span class="n">Value</span> <span class="n">Buffer</span><span class="p">:</span>
<span class="o">|</span><span class="n">Bytes</span> <span class="mi">0</span><span class="o">-</span><span class="mi">3</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">4</span><span class="o">-</span><span class="mi">7</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">8</span><span class="o">-</span><span class="mi">11</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">12</span><span class="o">-</span><span class="mi">15</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">16</span><span class="o">-</span><span class="mi">19</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">20</span><span class="o">-</span><span class="mi">63</span> <span class="o">|</span>
<span class="o">|------------|-------------|-------------|-------------|-------------|-------------|</span>
<span class="o">|</span> <span class="mi">1</span> <span class="o">|</span> <span class="n">unspecified</span> <span class="o">|</span> <span class="mi">2</span> <span class="o">|</span> <span class="mi">4</span> <span class="o">|</span> <span class="mi">8</span> <span class="o">|</span> <span class="n">unspecified</span> <span class="o">|</span>
</pre></div>
</div>
<p><strong>Example Layout: Non-null int32 Array</strong></p>
<p><code class="docutils literal notranslate"><span class="pre">[1,</span> <span class="pre">2,</span> <span class="pre">3,</span> <span class="pre">4,</span> <span class="pre">8]</span></code> has two possible layouts:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="o">*</span> <span class="n">Length</span><span class="p">:</span> <span class="mi">5</span><span class="p">,</span> <span class="n">Null</span> <span class="n">count</span><span class="p">:</span> <span class="mi">0</span>
<span class="o">*</span> <span class="n">Validity</span> <span class="n">bitmap</span> <span class="n">buffer</span><span class="p">:</span>
<span class="o">|</span> <span class="n">Byte</span> <span class="mi">0</span> <span class="p">(</span><span class="n">validity</span> <span class="n">bitmap</span><span class="p">)</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">1</span><span class="o">-</span><span class="mi">63</span> <span class="o">|</span>
<span class="o">|--------------------------|-----------------------|</span>
<span class="o">|</span> <span class="mi">00011111</span> <span class="o">|</span> <span class="mi">0</span> <span class="p">(</span><span class="n">padding</span><span class="p">)</span> <span class="o">|</span>
<span class="o">*</span> <span class="n">Value</span> <span class="n">Buffer</span><span class="p">:</span>
<span class="o">|</span><span class="n">Bytes</span> <span class="mi">0</span><span class="o">-</span><span class="mi">3</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">4</span><span class="o">-</span><span class="mi">7</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">8</span><span class="o">-</span><span class="mi">11</span> <span class="o">|</span> <span class="nb">bytes</span> <span class="mi">12</span><span class="o">-</span><span class="mi">15</span> <span class="o">|</span> <span class="nb">bytes</span> <span class="mi">16</span><span class="o">-</span><span class="mi">19</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">20</span><span class="o">-</span><span class="mi">63</span> <span class="o">|</span>
<span class="o">|------------|-------------|-------------|-------------|-------------|-------------|</span>
<span class="o">|</span> <span class="mi">1</span> <span class="o">|</span> <span class="mi">2</span> <span class="o">|</span> <span class="mi">3</span> <span class="o">|</span> <span class="mi">4</span> <span class="o">|</span> <span class="mi">8</span> <span class="o">|</span> <span class="n">unspecified</span> <span class="o">|</span>
</pre></div>
</div>
<p>or with the bitmap elided:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="o">*</span> <span class="n">Length</span> <span class="mi">5</span><span class="p">,</span> <span class="n">Null</span> <span class="n">count</span><span class="p">:</span> <span class="mi">0</span>
<span class="o">*</span> <span class="n">Validity</span> <span class="n">bitmap</span> <span class="n">buffer</span><span class="p">:</span> <span class="n">Not</span> <span class="n">required</span>
<span class="o">*</span> <span class="n">Value</span> <span class="n">Buffer</span><span class="p">:</span>
<span class="o">|</span><span class="n">Bytes</span> <span class="mi">0</span><span class="o">-</span><span class="mi">3</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">4</span><span class="o">-</span><span class="mi">7</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">8</span><span class="o">-</span><span class="mi">11</span> <span class="o">|</span> <span class="nb">bytes</span> <span class="mi">12</span><span class="o">-</span><span class="mi">15</span> <span class="o">|</span> <span class="nb">bytes</span> <span class="mi">16</span><span class="o">-</span><span class="mi">19</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">20</span><span class="o">-</span><span class="mi">63</span> <span class="o">|</span>
<span class="o">|------------|-------------|-------------|-------------|-------------|-------------|</span>
<span class="o">|</span> <span class="mi">1</span> <span class="o">|</span> <span class="mi">2</span> <span class="o">|</span> <span class="mi">3</span> <span class="o">|</span> <span class="mi">4</span> <span class="o">|</span> <span class="mi">8</span> <span class="o">|</span> <span class="n">unspecified</span> <span class="o">|</span>
</pre></div>
</div>
</section>
<section id="variable-size-binary-layout">
<h3>Variable-size Binary Layout<a class="headerlink" href="#variable-size-binary-layout" title="Permalink to this headline">#</a></h3>
<p>Each value in this layout consists of 0 or more bytes. While primitive
arrays have a single values buffer, variable-size binary have an
<strong>offsets</strong> buffer and <strong>data</strong> buffer.</p>
<p>The offsets buffer contains <cite>length + 1</cite> signed integers (either
32-bit or 64-bit, depending on the logical type), which encode the
start position of each slot in the data buffer. The length of the
value in each slot is computed using the difference between the offset
at that slot’s index and the subsequent offset. For example, the
position and length of slot j is computed as:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">slot_position</span> <span class="o">=</span> <span class="n">offsets</span><span class="p">[</span><span class="n">j</span><span class="p">]</span>
<span class="n">slot_length</span> <span class="o">=</span> <span class="n">offsets</span><span class="p">[</span><span class="n">j</span> <span class="o">+</span> <span class="mi">1</span><span class="p">]</span> <span class="o">-</span> <span class="n">offsets</span><span class="p">[</span><span class="n">j</span><span class="p">]</span> <span class="o">//</span> <span class="p">(</span><span class="k">for</span> <span class="mi">0</span> <span class="o">&lt;=</span> <span class="n">j</span> <span class="o">&lt;</span> <span class="n">length</span><span class="p">)</span>
</pre></div>
</div>
<p>It should be noted that a null value may have a positive slot length.
That is, a null value may occupy a <strong>non-empty</strong> memory space in the data
buffer. When this is true, the content of the corresponding memory space
is undefined.</p>
<p>Offsets must be monotonically increasing, that is <code class="docutils literal notranslate"><span class="pre">offsets[j+1]</span> <span class="pre">&gt;=</span> <span class="pre">offsets[j]</span></code>
for <code class="docutils literal notranslate"><span class="pre">0</span> <span class="pre">&lt;=</span> <span class="pre">j</span> <span class="pre">&lt;</span> <span class="pre">length</span></code>, even for null slots. This property ensures the
location for all values is valid and well defined.</p>
<p>Generally the first slot in the offsets array is 0, and the last slot
is the length of the values array. When serializing this layout, we
recommend normalizing the offsets to start at 0.</p>
</section>
<section id="variable-size-list-layout">
<h3>Variable-size List Layout<a class="headerlink" href="#variable-size-list-layout" title="Permalink to this headline">#</a></h3>
<p>List is a nested type which is semantically similar to variable-size
binary. It is defined by two buffers, a validity bitmap and an offsets
buffer, and a child array. The offsets are the same as in the
variable-size binary case, and both 32-bit and 64-bit signed integer
offsets are supported options for the offsets. Rather than referencing
an additional data buffer, instead these offsets reference the child
array.</p>
<p>Similar to the layout of variable-size binary, a null value may
correspond to a <strong>non-empty</strong> segment in the child array. When this is
true, the content of the corresponding segment can be arbitrary.</p>
<p>A list type is specified like <code class="docutils literal notranslate"><span class="pre">List&lt;T&gt;</span></code>, where <code class="docutils literal notranslate"><span class="pre">T</span></code> is any type
(primitive or nested). In these examples we use 32-bit offsets where
the 64-bit offset version would be denoted by <code class="docutils literal notranslate"><span class="pre">LargeList&lt;T&gt;</span></code>.</p>
<p><strong>Example Layout: ``List&lt;Int8&gt;`` Array</strong></p>
<p>We illustrate an example of <code class="docutils literal notranslate"><span class="pre">List&lt;Int8&gt;</span></code> with length 4 having values:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="p">[[</span><span class="mi">12</span><span class="p">,</span> <span class="o">-</span><span class="mi">7</span><span class="p">,</span> <span class="mi">25</span><span class="p">],</span> <span class="n">null</span><span class="p">,</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span> <span class="o">-</span><span class="mi">127</span><span class="p">,</span> <span class="mi">127</span><span class="p">,</span> <span class="mi">50</span><span class="p">],</span> <span class="p">[]]</span>
</pre></div>
</div>
<p>will have the following representation:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="o">*</span> <span class="n">Length</span><span class="p">:</span> <span class="mi">4</span><span class="p">,</span> <span class="n">Null</span> <span class="n">count</span><span class="p">:</span> <span class="mi">1</span>
<span class="o">*</span> <span class="n">Validity</span> <span class="n">bitmap</span> <span class="n">buffer</span><span class="p">:</span>
<span class="o">|</span> <span class="n">Byte</span> <span class="mi">0</span> <span class="p">(</span><span class="n">validity</span> <span class="n">bitmap</span><span class="p">)</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">1</span><span class="o">-</span><span class="mi">63</span> <span class="o">|</span>
<span class="o">|--------------------------|-----------------------|</span>
<span class="o">|</span> <span class="mi">00001101</span> <span class="o">|</span> <span class="mi">0</span> <span class="p">(</span><span class="n">padding</span><span class="p">)</span> <span class="o">|</span>
<span class="o">*</span> <span class="n">Offsets</span> <span class="n">buffer</span> <span class="p">(</span><span class="n">int32</span><span class="p">)</span>
<span class="o">|</span> <span class="n">Bytes</span> <span class="mi">0</span><span class="o">-</span><span class="mi">3</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">4</span><span class="o">-</span><span class="mi">7</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">8</span><span class="o">-</span><span class="mi">11</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">12</span><span class="o">-</span><span class="mi">15</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">16</span><span class="o">-</span><span class="mi">19</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">20</span><span class="o">-</span><span class="mi">63</span> <span class="o">|</span>
<span class="o">|------------|-------------|-------------|-------------|-------------|-------------|</span>
<span class="o">|</span> <span class="mi">0</span> <span class="o">|</span> <span class="mi">3</span> <span class="o">|</span> <span class="mi">3</span> <span class="o">|</span> <span class="mi">7</span> <span class="o">|</span> <span class="mi">7</span> <span class="o">|</span> <span class="n">unspecified</span> <span class="o">|</span>
<span class="o">*</span> <span class="n">Values</span> <span class="n">array</span> <span class="p">(</span><span class="n">Int8array</span><span class="p">):</span>
<span class="o">*</span> <span class="n">Length</span><span class="p">:</span> <span class="mi">7</span><span class="p">,</span> <span class="n">Null</span> <span class="n">count</span><span class="p">:</span> <span class="mi">0</span>
<span class="o">*</span> <span class="n">Validity</span> <span class="n">bitmap</span> <span class="n">buffer</span><span class="p">:</span> <span class="n">Not</span> <span class="n">required</span>
<span class="o">*</span> <span class="n">Values</span> <span class="n">buffer</span> <span class="p">(</span><span class="n">int8</span><span class="p">)</span>
<span class="o">|</span> <span class="n">Bytes</span> <span class="mi">0</span><span class="o">-</span><span class="mi">6</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">7</span><span class="o">-</span><span class="mi">63</span> <span class="o">|</span>
<span class="o">|------------------------------|-------------|</span>
<span class="o">|</span> <span class="mi">12</span><span class="p">,</span> <span class="o">-</span><span class="mi">7</span><span class="p">,</span> <span class="mi">25</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="o">-</span><span class="mi">127</span><span class="p">,</span> <span class="mi">127</span><span class="p">,</span> <span class="mi">50</span> <span class="o">|</span> <span class="n">unspecified</span> <span class="o">|</span>
</pre></div>
</div>
<p><strong>Example Layout: ``List&lt;List&lt;Int8&gt;&gt;``</strong></p>
<p><code class="docutils literal notranslate"><span class="pre">[[[1,</span> <span class="pre">2],</span> <span class="pre">[3,</span> <span class="pre">4]],</span> <span class="pre">[[5,</span> <span class="pre">6,</span> <span class="pre">7],</span> <span class="pre">null,</span> <span class="pre">[8]],</span> <span class="pre">[[9,</span> <span class="pre">10]]]</span></code></p>
<p>will be represented as follows:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span>* Length 3
* Nulls count: 0
* Validity bitmap buffer: Not required
* Offsets buffer (int32)
| Bytes 0-3 | Bytes 4-7 | Bytes 8-11 | Bytes 12-15 | Bytes 16-63 |
|------------|------------|------------|-------------|-------------|
| 0 | 2 | 5 | 6 | unspecified |
* Values array (`List&lt;Int8&gt;`)
* Length: 6, Null count: 1
* Validity bitmap buffer:
| Byte 0 (validity bitmap) | Bytes 1-63 |
|--------------------------|-------------|
| 00110111 | 0 (padding) |
* Offsets buffer (int32)
| Bytes 0-27 | Bytes 28-63 |
|----------------------|-------------|
| 0, 2, 4, 7, 7, 8, 10 | unspecified |
* Values array (Int8):
* Length: 10, Null count: 0
* Validity bitmap buffer: Not required
| Bytes 0-9 | Bytes 10-63 |
|-------------------------------|-------------|
| 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 | unspecified |
</pre></div>
</div>
</section>
<section id="fixed-size-list-layout">
<h3>Fixed-Size List Layout<a class="headerlink" href="#fixed-size-list-layout" title="Permalink to this headline">#</a></h3>
<p>Fixed-Size List is a nested type in which each array slot contains a
fixed-size sequence of values all having the same type.</p>
<p>A fixed size list type is specified like <code class="docutils literal notranslate"><span class="pre">FixedSizeList&lt;T&gt;[N]</span></code>,
where <code class="docutils literal notranslate"><span class="pre">T</span></code> is any type (primitive or nested) and <code class="docutils literal notranslate"><span class="pre">N</span></code> is a 32-bit
signed integer representing the length of the lists.</p>
<p>A fixed size list array is represented by a values array, which is a
child array of type T. T may also be a nested type. The value in slot
<code class="docutils literal notranslate"><span class="pre">j</span></code> of a fixed size list array is stored in an <code class="docutils literal notranslate"><span class="pre">N</span></code>-long slice of
the values array, starting at an offset of <code class="docutils literal notranslate"><span class="pre">j</span> <span class="pre">*</span> <span class="pre">N</span></code>.</p>
<p><strong>Example Layout: ``FixedSizeList&lt;byte&gt;[4]`` Array</strong></p>
<p>Here we illustrate <code class="docutils literal notranslate"><span class="pre">FixedSizeList&lt;byte&gt;[4]</span></code>.</p>
<p>For an array of length 4 with respective values:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="p">[[</span><span class="mi">192</span><span class="p">,</span> <span class="mi">168</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">12</span><span class="p">],</span> <span class="n">null</span><span class="p">,</span> <span class="p">[</span><span class="mi">192</span><span class="p">,</span> <span class="mi">168</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">25</span><span class="p">],</span> <span class="p">[</span><span class="mi">192</span><span class="p">,</span> <span class="mi">168</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">]]</span>
</pre></div>
</div>
<p>will have the following representation:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="o">*</span> <span class="n">Length</span><span class="p">:</span> <span class="mi">4</span><span class="p">,</span> <span class="n">Null</span> <span class="n">count</span><span class="p">:</span> <span class="mi">1</span>
<span class="o">*</span> <span class="n">Validity</span> <span class="n">bitmap</span> <span class="n">buffer</span><span class="p">:</span>
<span class="o">|</span> <span class="n">Byte</span> <span class="mi">0</span> <span class="p">(</span><span class="n">validity</span> <span class="n">bitmap</span><span class="p">)</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">1</span><span class="o">-</span><span class="mi">63</span> <span class="o">|</span>
<span class="o">|--------------------------|-----------------------|</span>
<span class="o">|</span> <span class="mi">00001101</span> <span class="o">|</span> <span class="mi">0</span> <span class="p">(</span><span class="n">padding</span><span class="p">)</span> <span class="o">|</span>
<span class="o">*</span> <span class="n">Values</span> <span class="n">array</span> <span class="p">(</span><span class="n">byte</span> <span class="n">array</span><span class="p">):</span>
<span class="o">*</span> <span class="n">Length</span><span class="p">:</span> <span class="mi">16</span><span class="p">,</span> <span class="n">Null</span> <span class="n">count</span><span class="p">:</span> <span class="mi">0</span>
<span class="o">*</span> <span class="n">validity</span> <span class="n">bitmap</span> <span class="n">buffer</span><span class="p">:</span> <span class="n">Not</span> <span class="n">required</span>
<span class="o">|</span> <span class="n">Bytes</span> <span class="mi">0</span><span class="o">-</span><span class="mi">3</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">4</span><span class="o">-</span><span class="mi">7</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">8</span><span class="o">-</span><span class="mi">15</span> <span class="o">|</span>
<span class="o">|-----------------|-------------|---------------------------------|</span>
<span class="o">|</span> <span class="mi">192</span><span class="p">,</span> <span class="mi">168</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">12</span> <span class="o">|</span> <span class="n">unspecified</span> <span class="o">|</span> <span class="mi">192</span><span class="p">,</span> <span class="mi">168</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">25</span><span class="p">,</span> <span class="mi">192</span><span class="p">,</span> <span class="mi">168</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">1</span> <span class="o">|</span>
</pre></div>
</div>
</section>
<section id="struct-layout">
<h3>Struct Layout<a class="headerlink" href="#struct-layout" title="Permalink to this headline">#</a></h3>
<p>A struct is a nested type parameterized by an ordered sequence of
types (which can all be distinct), called its fields. Each field must
have a UTF8-encoded name, and these field names are part of the type
metadata.</p>
<p>Physically, a struct array has one child array for each field. The
child arrays are independent and need not be adjacent to each other in
memory. A struct array also has a validity bitmap to encode top-level
validity information.</p>
<p>For example, the struct (field names shown here as strings for illustration
purposes):</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">Struct</span> <span class="o">&lt;</span>
<span class="n">name</span><span class="p">:</span> <span class="n">VarBinary</span>
<span class="n">age</span><span class="p">:</span> <span class="n">Int32</span>
<span class="o">&gt;</span>
</pre></div>
</div>
<p>has two child arrays, one <code class="docutils literal notranslate"><span class="pre">VarBinary</span></code> array (using variable-size binary
layout) and one 4-byte primitive value array having <code class="docutils literal notranslate"><span class="pre">Int32</span></code> logical
type.</p>
<p><strong>Example Layout: ``Struct&lt;VarBinary, Int32&gt;``</strong></p>
<p>The layout for <code class="docutils literal notranslate"><span class="pre">[{'joe',</span> <span class="pre">1},</span> <span class="pre">{null,</span> <span class="pre">2},</span> <span class="pre">null,</span> <span class="pre">{'mark',</span> <span class="pre">4}]</span></code> would be:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span>* Length: 4, Null count: 1
* Validity bitmap buffer:
|Byte 0 (validity bitmap) | Bytes 1-63 |
|-------------------------|-----------------------|
| 00001011 | 0 (padding) |
* Children arrays:
* field-0 array (`VarBinary`):
* Length: 4, Null count: 2
* Validity bitmap buffer:
| Byte 0 (validity bitmap) | Bytes 1-63 |
|--------------------------|-----------------------|
| 00001001 | 0 (padding) |
* Offsets buffer:
| Bytes 0-19 |
|----------------|
| 0, 3, 3, 3, 7 |
* Values array:
* Length: 7, Null count: 0
* Validity bitmap buffer: Not required
* Value buffer:
| Bytes 0-6 |
|----------------|
| joemark |
* field-1 array (int32 array):
* Length: 4, Null count: 1
* Validity bitmap buffer:
| Byte 0 (validity bitmap) | Bytes 1-63 |
|--------------------------|-----------------------|
| 00001011 | 0 (padding) |
* Value Buffer:
|Bytes 0-3 | Bytes 4-7 | Bytes 8-11 | Bytes 12-15 | Bytes 16-63 |
|------------|-------------|-------------|-------------|-------------|
| 1 | 2 | unspecified | 4 | unspecified |
</pre></div>
</div>
<section id="struct-validity">
<h4>Struct Validity<a class="headerlink" href="#struct-validity" title="Permalink to this headline">#</a></h4>
<p>A struct array has its own validity bitmap that is independent of its
child arrays’ validity bitmaps. The validity bitmap for the struct
array might indicate a null when one or more of its child arrays has
a non-null value in its corresponding slot; or conversely, a child
array might have a null in its validity bitmap while the struct array’s
validity bitmap shows a non-null value.</p>
<p>Therefore, to know whether a particular child entry is valid, one must
take the logical AND of the corresponding bits in the two validity bitmaps
(the struct array’s and the child array’s).</p>
<p>This is illustrated in the example above, the child arrays have valid entries
for the null struct but they are “hidden” by the struct array’s validity
bitmap. However, when treated independently, corresponding entries of the
children array will be non-null.</p>
</section>
</section>
<section id="union-layout">
<h3>Union Layout<a class="headerlink" href="#union-layout" title="Permalink to this headline">#</a></h3>
<p>A union is defined by an ordered sequence of types; each slot in the
union can have a value chosen from these types. The types are named
like a struct’s fields, and the names are part of the type metadata.</p>
<p>Unlike other data types, unions do not have their own validity bitmap. Instead,
the nullness of each slot is determined exclusively by the child arrays which
are composed to create the union.</p>
<p>We define two distinct union types, “dense” and “sparse”, that are
optimized for different use cases.</p>
<section id="dense-union">
<h4>Dense Union<a class="headerlink" href="#dense-union" title="Permalink to this headline">#</a></h4>
<p>Dense union represents a mixed-type array with 5 bytes of overhead for
each value. Its physical layout is as follows:</p>
<ul class="simple">
<li><p>One child array for each type</p></li>
<li><p>Types buffer: A buffer of 8-bit signed integers. Each type in the
union has a corresponding type id whose values are found in this
buffer. A union with more than 127 possible types can be modeled as
a union of unions.</p></li>
<li><p>Offsets buffer: A buffer of signed Int32 values indicating the
relative offset into the respective child array for the type in a
given slot. The respective offsets for each child value array must
be in order / increasing.</p></li>
</ul>
<p><strong>Example Layout: ``DenseUnion&lt;f: Float32, i: Int32&gt;``</strong></p>
<p>For the union array:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="p">[{</span><span class="n">f</span><span class="o">=</span><span class="mf">1.2</span><span class="p">},</span> <span class="n">null</span><span class="p">,</span> <span class="p">{</span><span class="n">f</span><span class="o">=</span><span class="mf">3.4</span><span class="p">},</span> <span class="p">{</span><span class="n">i</span><span class="o">=</span><span class="mi">5</span><span class="p">}]</span>
</pre></div>
</div>
<p>will have the following layout:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="o">*</span> <span class="n">Length</span><span class="p">:</span> <span class="mi">4</span><span class="p">,</span> <span class="n">Null</span> <span class="n">count</span><span class="p">:</span> <span class="mi">0</span>
<span class="o">*</span> <span class="n">Types</span> <span class="n">buffer</span><span class="p">:</span>
<span class="o">|</span><span class="n">Byte</span> <span class="mi">0</span> <span class="o">|</span> <span class="n">Byte</span> <span class="mi">1</span> <span class="o">|</span> <span class="n">Byte</span> <span class="mi">2</span> <span class="o">|</span> <span class="n">Byte</span> <span class="mi">3</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">4</span><span class="o">-</span><span class="mi">63</span> <span class="o">|</span>
<span class="o">|---------|-------------|----------|----------|-------------|</span>
<span class="o">|</span> <span class="mi">0</span> <span class="o">|</span> <span class="mi">0</span> <span class="o">|</span> <span class="mi">0</span> <span class="o">|</span> <span class="mi">1</span> <span class="o">|</span> <span class="n">unspecified</span> <span class="o">|</span>
<span class="o">*</span> <span class="n">Offset</span> <span class="n">buffer</span><span class="p">:</span>
<span class="o">|</span><span class="n">Bytes</span> <span class="mi">0</span><span class="o">-</span><span class="mi">3</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">4</span><span class="o">-</span><span class="mi">7</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">8</span><span class="o">-</span><span class="mi">11</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">12</span><span class="o">-</span><span class="mi">15</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">16</span><span class="o">-</span><span class="mi">63</span> <span class="o">|</span>
<span class="o">|----------|-------------|------------|-------------|-------------|</span>
<span class="o">|</span> <span class="mi">0</span> <span class="o">|</span> <span class="mi">1</span> <span class="o">|</span> <span class="mi">2</span> <span class="o">|</span> <span class="mi">0</span> <span class="o">|</span> <span class="n">unspecified</span> <span class="o">|</span>
<span class="o">*</span> <span class="n">Children</span> <span class="n">arrays</span><span class="p">:</span>
<span class="o">*</span> <span class="n">Field</span><span class="o">-</span><span class="mi">0</span> <span class="n">array</span> <span class="p">(</span><span class="n">f</span><span class="p">:</span> <span class="n">Float32</span><span class="p">):</span>
<span class="o">*</span> <span class="n">Length</span><span class="p">:</span> <span class="mi">2</span><span class="p">,</span> <span class="n">Null</span> <span class="n">count</span><span class="p">:</span> <span class="mi">1</span>
<span class="o">*</span> <span class="n">Validity</span> <span class="n">bitmap</span> <span class="n">buffer</span><span class="p">:</span> <span class="mi">00000101</span>
<span class="o">*</span> <span class="n">Value</span> <span class="n">Buffer</span><span class="p">:</span>
<span class="o">|</span> <span class="n">Bytes</span> <span class="mi">0</span><span class="o">-</span><span class="mi">11</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">12</span><span class="o">-</span><span class="mi">63</span> <span class="o">|</span>
<span class="o">|----------------|-------------|</span>
<span class="o">|</span> <span class="mf">1.2</span><span class="p">,</span> <span class="n">null</span><span class="p">,</span> <span class="mf">3.4</span> <span class="o">|</span> <span class="n">unspecified</span> <span class="o">|</span>
<span class="o">*</span> <span class="n">Field</span><span class="o">-</span><span class="mi">1</span> <span class="n">array</span> <span class="p">(</span><span class="n">i</span><span class="p">:</span> <span class="n">Int32</span><span class="p">):</span>
<span class="o">*</span> <span class="n">Length</span><span class="p">:</span> <span class="mi">1</span><span class="p">,</span> <span class="n">Null</span> <span class="n">count</span><span class="p">:</span> <span class="mi">0</span>
<span class="o">*</span> <span class="n">Validity</span> <span class="n">bitmap</span> <span class="n">buffer</span><span class="p">:</span> <span class="n">Not</span> <span class="n">required</span>
<span class="o">*</span> <span class="n">Value</span> <span class="n">Buffer</span><span class="p">:</span>
<span class="o">|</span> <span class="n">Bytes</span> <span class="mi">0</span><span class="o">-</span><span class="mi">3</span> <span class="o">|</span> <span class="n">Bytes</span> <span class="mi">4</span><span class="o">-</span><span class="mi">63</span> <span class="o">|</span>
<span class="o">|-----------|-------------|</span>
<span class="o">|</span> <span class="mi">5</span> <span class="o">|</span> <span class="n">unspecified</span> <span class="o">|</span>
</pre></div>
</div>
</section>
<section id="sparse-union">
<h4>Sparse Union<a class="headerlink" href="#sparse-union" title="Permalink to this headline">#</a></h4>
<p>A sparse union has the same structure as a dense union, with the omission of
the offsets array. In this case, the child arrays are each equal in length to
the length of the union.</p>
<p>While a sparse union may use significantly more space compared with a
dense union, it has some advantages that may be desirable in certain
use cases:</p>
<ul class="simple">
<li><p>A sparse union is more amenable to vectorized expression evaluation in some use cases.</p></li>
<li><p>Equal-length arrays can be interpreted as a union by only defining the types array.</p></li>
</ul>
<p><strong>Example layout: ``SparseUnion&lt;i: Int32, f: Float32, s: VarBinary&gt;``</strong></p>
<p>For the union array:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="p">[{</span><span class="n">i</span><span class="o">=</span><span class="mi">5</span><span class="p">},</span> <span class="p">{</span><span class="n">f</span><span class="o">=</span><span class="mf">1.2</span><span class="p">},</span> <span class="p">{</span><span class="n">s</span><span class="o">=</span><span class="s1">&#39;joe&#39;</span><span class="p">},</span> <span class="p">{</span><span class="n">f</span><span class="o">=</span><span class="mf">3.4</span><span class="p">},</span> <span class="p">{</span><span class="n">i</span><span class="o">=</span><span class="mi">4</span><span class="p">},</span> <span class="p">{</span><span class="n">s</span><span class="o">=</span><span class="s1">&#39;mark&#39;</span><span class="p">}]</span>
</pre></div>
</div>
<p>will have the following layout:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span>* Length: 6, Null count: 0
* Types buffer:
| Byte 0 | Byte 1 | Byte 2 | Byte 3 | Byte 4 | Byte 5 | Bytes 6-63 |
|------------|-------------|-------------|-------------|-------------|--------------|-----------------------|
| 0 | 1 | 2 | 1 | 0 | 2 | unspecified (padding) |
* Children arrays:
* i (Int32):
* Length: 6, Null count: 4
* Validity bitmap buffer:
|Byte 0 (validity bitmap) | Bytes 1-63 |
|-------------------------|-----------------------|
|00010001 | 0 (padding) |
* Value buffer:
|Bytes 0-3 | Bytes 4-7 | Bytes 8-11 | Bytes 12-15 | Bytes 16-19 | Bytes 20-23 | Bytes 24-63 |
|------------|-------------|-------------|-------------|-------------|--------------|-----------------------|
| 5 | unspecified | unspecified | unspecified | 4 | unspecified | unspecified (padding) |
* f (Float32):
* Length: 6, Null count: 4
* Validity bitmap buffer:
|Byte 0 (validity bitmap) | Bytes 1-63 |
|-------------------------|-----------------------|
| 00001010 | 0 (padding) |
* Value buffer:
|Bytes 0-3 | Bytes 4-7 | Bytes 8-11 | Bytes 12-15 | Bytes 16-19 | Bytes 20-23 | Bytes 24-63 |
|-------------|-------------|-------------|-------------|-------------|--------------|-----------------------|
| unspecified | 1.2 | unspecified | 3.4 | unspecified | unspecified | unspecified (padding) |
* s (`VarBinary`)
* Length: 6, Null count: 4
* Validity bitmap buffer:
| Byte 0 (validity bitmap) | Bytes 1-63 |
|--------------------------|-----------------------|
| 00100100 | 0 (padding) |
* Offsets buffer (Int32)
| Bytes 0-3 | Bytes 4-7 | Bytes 8-11 | Bytes 12-15 | Bytes 16-19 | Bytes 20-23 | Bytes 24-27 | Bytes 28-63 |
|------------|-------------|-------------|-------------|-------------|-------------|-------------|-------------|
| 0 | 0 | 0 | 3 | 3 | 3 | 7 | unspecified |
* Values array (VarBinary):
* Length: 7, Null count: 0
* Validity bitmap buffer: Not required
| Bytes 0-6 | Bytes 7-63 |
|------------|-----------------------|
| joemark | unspecified (padding) |
</pre></div>
</div>
<p>Only the slot in the array corresponding to the type index is considered. All
“unselected” values are ignored and could be any semantically correct array
value.</p>
</section>
</section>
<section id="null-layout">
<h3>Null Layout<a class="headerlink" href="#null-layout" title="Permalink to this headline">#</a></h3>
<p>We provide a simplified memory-efficient layout for the Null data type
where all values are null. In this case no memory buffers are
allocated.</p>
</section>
<section id="dictionary-encoded-layout">
<span id="id1"></span><h3>Dictionary-encoded Layout<a class="headerlink" href="#dictionary-encoded-layout" title="Permalink to this headline">#</a></h3>
<p>Dictionary encoding is a data representation technique to represent
values by integers referencing a <strong>dictionary</strong> usually consisting of
unique values. It can be effective when you have data with many
repeated values.</p>
<p>Any array can be dictionary-encoded. The dictionary is stored as an optional
property of an array. When a field is dictionary encoded, the values are
represented by an array of non-negative integers representing the index of the
value in the dictionary. The memory layout for a dictionary-encoded array is
the same as that of a primitive integer layout. The dictionary is handled as a
separate columnar array with its own respective layout.</p>
<p>As an example, you could have the following data:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="nb">type</span><span class="p">:</span> <span class="n">VarBinary</span>
<span class="p">[</span><span class="s1">&#39;foo&#39;</span><span class="p">,</span> <span class="s1">&#39;bar&#39;</span><span class="p">,</span> <span class="s1">&#39;foo&#39;</span><span class="p">,</span> <span class="s1">&#39;bar&#39;</span><span class="p">,</span> <span class="n">null</span><span class="p">,</span> <span class="s1">&#39;baz&#39;</span><span class="p">]</span>
</pre></div>
</div>
<p>In dictionary-encoded form, this could appear as:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">data</span> <span class="n">VarBinary</span> <span class="p">(</span><span class="n">dictionary</span><span class="o">-</span><span class="n">encoded</span><span class="p">)</span>
<span class="n">index_type</span><span class="p">:</span> <span class="n">Int32</span>
<span class="n">values</span><span class="p">:</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="n">null</span><span class="p">,</span> <span class="mi">2</span><span class="p">]</span>
<span class="n">dictionary</span>
<span class="nb">type</span><span class="p">:</span> <span class="n">VarBinary</span>
<span class="n">values</span><span class="p">:</span> <span class="p">[</span><span class="s1">&#39;foo&#39;</span><span class="p">,</span> <span class="s1">&#39;bar&#39;</span><span class="p">,</span> <span class="s1">&#39;baz&#39;</span><span class="p">]</span>
</pre></div>
</div>
<p>Note that a dictionary is permitted to contain duplicate values or
nulls:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">data</span> <span class="n">VarBinary</span> <span class="p">(</span><span class="n">dictionary</span><span class="o">-</span><span class="n">encoded</span><span class="p">)</span>
<span class="n">index_type</span><span class="p">:</span> <span class="n">Int32</span>
<span class="n">values</span><span class="p">:</span> <span class="p">[</span><span class="mi">0</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">3</span><span class="p">,</span> <span class="mi">1</span><span class="p">,</span> <span class="mi">4</span><span class="p">,</span> <span class="mi">2</span><span class="p">]</span>
<span class="n">dictionary</span>
<span class="nb">type</span><span class="p">:</span> <span class="n">VarBinary</span>
<span class="n">values</span><span class="p">:</span> <span class="p">[</span><span class="s1">&#39;foo&#39;</span><span class="p">,</span> <span class="s1">&#39;bar&#39;</span><span class="p">,</span> <span class="s1">&#39;baz&#39;</span><span class="p">,</span> <span class="s1">&#39;foo&#39;</span><span class="p">,</span> <span class="n">null</span><span class="p">]</span>
</pre></div>
</div>
<p>The null count of such arrays is dictated only by the validity bitmap
of its indices, irrespective of any null values in the dictionary.</p>
<p>Since unsigned integers can be more difficult to work with in some cases
(e.g. in the JVM), we recommend preferring signed integers over unsigned
integers for representing dictionary indices. Additionally, we recommend
avoiding using 64-bit unsigned integer indices unless they are required by an
application.</p>
<p>We discuss dictionary encoding as it relates to serialization further
below.</p>
</section>
<section id="buffer-listing-for-each-layout">
<h3>Buffer Listing for Each Layout<a class="headerlink" href="#buffer-listing-for-each-layout" title="Permalink to this headline">#</a></h3>
<p>For the avoidance of ambiguity, we provide listing the order and type
of memory buffers for each layout.</p>
<table class="colwidths-given table" id="id2">
<caption><span class="caption-text">Buffer Layouts</span><a class="headerlink" href="#id2" title="Permalink to this table">#</a></caption>
<colgroup>
<col style="width: 33%" />
<col style="width: 22%" />
<col style="width: 22%" />
<col style="width: 22%" />
</colgroup>
<thead>
<tr class="row-odd"><th class="head"><p>Layout Type</p></th>
<th class="head"><p>Buffer 0</p></th>
<th class="head"><p>Buffer 1</p></th>
<th class="head"><p>Buffer 2</p></th>
</tr>
</thead>
<tbody>
<tr class="row-even"><td><p>Primitive</p></td>
<td><p>validity</p></td>
<td><p>data</p></td>
<td></td>
</tr>
<tr class="row-odd"><td><p>Variable Binary</p></td>
<td><p>validity</p></td>
<td><p>offsets</p></td>
<td><p>data</p></td>
</tr>
<tr class="row-even"><td><p>List</p></td>
<td><p>validity</p></td>
<td><p>offsets</p></td>
<td></td>
</tr>
<tr class="row-odd"><td><p>Fixed-size List</p></td>
<td><p>validity</p></td>
<td></td>
<td></td>
</tr>
<tr class="row-even"><td><p>Struct</p></td>
<td><p>validity</p></td>
<td></td>
<td></td>
</tr>
<tr class="row-odd"><td><p>Sparse Union</p></td>
<td><p>type ids</p></td>
<td></td>
<td></td>
</tr>
<tr class="row-even"><td><p>Dense Union</p></td>
<td><p>type ids</p></td>
<td><p>offsets</p></td>
<td></td>
</tr>
<tr class="row-odd"><td><p>Null</p></td>
<td></td>
<td></td>
<td></td>
</tr>
<tr class="row-even"><td><p>Dictionary-encoded</p></td>
<td><p>validity</p></td>
<td><p>data (indices)</p></td>
<td></td>
</tr>
</tbody>
</table>
</section>
</section>
<section id="logical-types">
<h2>Logical Types<a class="headerlink" href="#logical-types" title="Permalink to this headline">#</a></h2>
<p>The <a class="reference external" href="https://github.com/apache/arrow/blob/master/format/Schema.fbs">Schema.fbs</a> defines built-in logical types supported by the
Arrow columnar format. Each logical type uses one of the above
physical layouts. Nested logical types may have different physical
layouts depending on the particular realization of the type.</p>
<p>We do not go into detail about the logical types definitions in this
document as we consider <a class="reference external" href="https://github.com/apache/arrow/blob/master/format/Schema.fbs">Schema.fbs</a> to be authoritative.</p>
</section>
<section id="serialization-and-interprocess-communication-ipc">
<span id="format-ipc"></span><h2>Serialization and Interprocess Communication (IPC)<a class="headerlink" href="#serialization-and-interprocess-communication-ipc" title="Permalink to this headline">#</a></h2>
<p>The primitive unit of serialized data in the columnar format is the
“record batch”. Semantically, a record batch is an ordered collection
of arrays, known as its <strong>fields</strong>, each having the same length as one
another but potentially different data types. A record batch’s field
names and types collectively form the batch’s <strong>schema</strong>.</p>
<p>In this section we define a protocol for serializing record batches
into a stream of binary payloads and reconstructing record batches
from these payloads without need for memory copying.</p>
<p>The columnar IPC protocol utilizes a one-way stream of binary messages
of these types:</p>
<ul class="simple">
<li><p>Schema</p></li>
<li><p>RecordBatch</p></li>
<li><p>DictionaryBatch</p></li>
</ul>
<p>We specify a so-called <em>encapsulated IPC message</em> format which
includes a serialized Flatbuffer type along with an optional message
body. We define this message format before describing how to serialize
each constituent IPC message type.</p>
<section id="encapsulated-message-format">
<h3>Encapsulated message format<a class="headerlink" href="#encapsulated-message-format" title="Permalink to this headline">#</a></h3>
<p>For simple streaming and file-based serialization, we define a
“encapsulated” message format for interprocess communication. Such
messages can be “deserialized” into in-memory Arrow array objects by
examining only the message metadata without any need to copy or move
any of the actual data.</p>
<p>The encapsulated binary message format is as follows:</p>
<ul class="simple">
<li><p>A 32-bit continuation indicator. The value <code class="docutils literal notranslate"><span class="pre">0xFFFFFFFF</span></code> indicates
a valid message. This component was introduced in version 0.15.0 in
part to address the 8-byte alignment requirement of Flatbuffers</p></li>
<li><p>A 32-bit little-endian length prefix indicating the metadata size</p></li>
<li><p>The message metadata as using the <code class="docutils literal notranslate"><span class="pre">Message</span></code> type defined in
<a class="reference external" href="https://github.com/apache/arrow/blob/master/format/Message.fbs">Message.fbs</a></p></li>
<li><p>Padding bytes to an 8-byte boundary</p></li>
<li><p>The message body, whose length must be a multiple of 8 bytes</p></li>
</ul>
<p>Schematically, we have:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="o">&lt;</span><span class="n">continuation</span><span class="p">:</span> <span class="mh">0xFFFFFFFF</span><span class="o">&gt;</span>
<span class="o">&lt;</span><span class="n">metadata_size</span><span class="p">:</span> <span class="n">int32</span><span class="o">&gt;</span>
<span class="o">&lt;</span><span class="n">metadata_flatbuffer</span><span class="p">:</span> <span class="nb">bytes</span><span class="o">&gt;</span>
<span class="o">&lt;</span><span class="n">padding</span><span class="o">&gt;</span>
<span class="o">&lt;</span><span class="n">message</span> <span class="n">body</span><span class="o">&gt;</span>
</pre></div>
</div>
<p>The complete serialized message must be a multiple of 8 bytes so that messages
can be relocated between streams. Otherwise the amount of padding between the
metadata and the message body could be non-deterministic.</p>
<p>The <code class="docutils literal notranslate"><span class="pre">metadata_size</span></code> includes the size of the <code class="docutils literal notranslate"><span class="pre">Message</span></code> plus
padding. The <code class="docutils literal notranslate"><span class="pre">metadata_flatbuffer</span></code> contains a serialized <code class="docutils literal notranslate"><span class="pre">Message</span></code>
Flatbuffer value, which internally includes:</p>
<ul class="simple">
<li><p>A version number</p></li>
<li><p>A particular message value (one of <code class="docutils literal notranslate"><span class="pre">Schema</span></code>, <code class="docutils literal notranslate"><span class="pre">RecordBatch</span></code>, or
<code class="docutils literal notranslate"><span class="pre">DictionaryBatch</span></code>)</p></li>
<li><p>The size of the message body</p></li>
<li><p>A <code class="docutils literal notranslate"><span class="pre">custom_metadata</span></code> field for any application-supplied metadata</p></li>
</ul>
<p>When read from an input stream, generally the <code class="docutils literal notranslate"><span class="pre">Message</span></code> metadata is
initially parsed and validated to obtain the body size. Then the body
can be read.</p>
</section>
<section id="schema-message">
<h3>Schema message<a class="headerlink" href="#schema-message" title="Permalink to this headline">#</a></h3>
<p>The Flatbuffers files <a class="reference external" href="https://github.com/apache/arrow/blob/master/format/Schema.fbs">Schema.fbs</a> contains the definitions for all
built-in logical data types and the <code class="docutils literal notranslate"><span class="pre">Schema</span></code> metadata type which
represents the schema of a given record batch. A schema consists of
an ordered sequence of fields, each having a name and type. A
serialized <code class="docutils literal notranslate"><span class="pre">Schema</span></code> does not contain any data buffers, only type
metadata.</p>
<p>The <code class="docutils literal notranslate"><span class="pre">Field</span></code> Flatbuffers type contains the metadata for a single
array. This includes:</p>
<ul class="simple">
<li><p>The field’s name</p></li>
<li><p>The field’s logical type</p></li>
<li><p>Whether the field is semantically nullable. While this has no
bearing on the array’s physical layout, many systems distinguish
nullable and non-nullable fields and we want to allow them to
preserve this metadata to enable faithful schema round trips.</p></li>
<li><p>A collection of child <code class="docutils literal notranslate"><span class="pre">Field</span></code> values, for nested types</p></li>
<li><p>A <code class="docutils literal notranslate"><span class="pre">dictionary</span></code> property indicating whether the field is
dictionary-encoded or not. If it is, a dictionary “id” is assigned
to allow matching a subsequent dictionary IPC message with the
appropriate field.</p></li>
</ul>
<p>We additionally provide both schema-level and field-level
<code class="docutils literal notranslate"><span class="pre">custom_metadata</span></code> attributes allowing for systems to insert their
own application defined metadata to customize behavior.</p>
</section>
<section id="recordbatch-message">
<h3>RecordBatch message<a class="headerlink" href="#recordbatch-message" title="Permalink to this headline">#</a></h3>
<p>A RecordBatch message contains the actual data buffers corresponding
to the physical memory layout determined by a schema. The metadata for
this message provides the location and size of each buffer, permitting
Array data structures to be reconstructed using pointer arithmetic and
thus no memory copying.</p>
<p>The serialized form of the record batch is the following:</p>
<ul class="simple">
<li><p>The <code class="docutils literal notranslate"><span class="pre">data</span> <span class="pre">header</span></code>, defined as the <code class="docutils literal notranslate"><span class="pre">RecordBatch</span></code> type in
<a class="reference external" href="https://github.com/apache/arrow/blob/master/format/Message.fbs">Message.fbs</a>.</p></li>
<li><p>The <code class="docutils literal notranslate"><span class="pre">body</span></code>, a flat sequence of memory buffers written end-to-end
with appropriate padding to ensure a minimum of 8-byte alignment</p></li>
</ul>
<p>The data header contains the following:</p>
<ul class="simple">
<li><p>The length and null count for each flattened field in the record
batch</p></li>
<li><p>The memory offset and length of each constituent <code class="docutils literal notranslate"><span class="pre">Buffer</span></code> in the
record batch’s body</p></li>
</ul>
<p>Fields and buffers are flattened by a pre-order depth-first traversal
of the fields in the record batch. For example, let’s consider the
schema</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">col1</span><span class="p">:</span> <span class="n">Struct</span><span class="o">&lt;</span><span class="n">a</span><span class="p">:</span> <span class="n">Int32</span><span class="p">,</span> <span class="n">b</span><span class="p">:</span> <span class="n">List</span><span class="o">&lt;</span><span class="n">item</span><span class="p">:</span> <span class="n">Int64</span><span class="o">&gt;</span><span class="p">,</span> <span class="n">c</span><span class="p">:</span> <span class="n">Float64</span><span class="o">&gt;</span>
<span class="n">col2</span><span class="p">:</span> <span class="n">Utf8</span>
</pre></div>
</div>
<p>The flattened version of this is:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">FieldNode</span> <span class="mi">0</span><span class="p">:</span> <span class="n">Struct</span> <span class="n">name</span><span class="o">=</span><span class="s1">&#39;col1&#39;</span>
<span class="n">FieldNode</span> <span class="mi">1</span><span class="p">:</span> <span class="n">Int32</span> <span class="n">name</span><span class="o">=</span><span class="s1">&#39;a&#39;</span>
<span class="n">FieldNode</span> <span class="mi">2</span><span class="p">:</span> <span class="n">List</span> <span class="n">name</span><span class="o">=</span><span class="s1">&#39;b&#39;</span>
<span class="n">FieldNode</span> <span class="mi">3</span><span class="p">:</span> <span class="n">Int64</span> <span class="n">name</span><span class="o">=</span><span class="s1">&#39;item&#39;</span>
<span class="n">FieldNode</span> <span class="mi">4</span><span class="p">:</span> <span class="n">Float64</span> <span class="n">name</span><span class="o">=</span><span class="s1">&#39;c&#39;</span>
<span class="n">FieldNode</span> <span class="mi">5</span><span class="p">:</span> <span class="n">Utf8</span> <span class="n">name</span><span class="o">=</span><span class="s1">&#39;col2&#39;</span>
</pre></div>
</div>
<p>For the buffers produced, we would have the following (refer to the
table above):</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">buffer</span> <span class="mi">0</span><span class="p">:</span> <span class="n">field</span> <span class="mi">0</span> <span class="n">validity</span>
<span class="n">buffer</span> <span class="mi">1</span><span class="p">:</span> <span class="n">field</span> <span class="mi">1</span> <span class="n">validity</span>
<span class="n">buffer</span> <span class="mi">2</span><span class="p">:</span> <span class="n">field</span> <span class="mi">1</span> <span class="n">values</span>
<span class="n">buffer</span> <span class="mi">3</span><span class="p">:</span> <span class="n">field</span> <span class="mi">2</span> <span class="n">validity</span>
<span class="n">buffer</span> <span class="mi">4</span><span class="p">:</span> <span class="n">field</span> <span class="mi">2</span> <span class="n">offsets</span>
<span class="n">buffer</span> <span class="mi">5</span><span class="p">:</span> <span class="n">field</span> <span class="mi">3</span> <span class="n">validity</span>
<span class="n">buffer</span> <span class="mi">6</span><span class="p">:</span> <span class="n">field</span> <span class="mi">3</span> <span class="n">values</span>
<span class="n">buffer</span> <span class="mi">7</span><span class="p">:</span> <span class="n">field</span> <span class="mi">4</span> <span class="n">validity</span>
<span class="n">buffer</span> <span class="mi">8</span><span class="p">:</span> <span class="n">field</span> <span class="mi">4</span> <span class="n">values</span>
<span class="n">buffer</span> <span class="mi">9</span><span class="p">:</span> <span class="n">field</span> <span class="mi">5</span> <span class="n">validity</span>
<span class="n">buffer</span> <span class="mi">10</span><span class="p">:</span> <span class="n">field</span> <span class="mi">5</span> <span class="n">offsets</span>
<span class="n">buffer</span> <span class="mi">11</span><span class="p">:</span> <span class="n">field</span> <span class="mi">5</span> <span class="n">data</span>
</pre></div>
</div>
<p>The <code class="docutils literal notranslate"><span class="pre">Buffer</span></code> Flatbuffers value describes the location and size of a
piece of memory. Generally these are interpreted relative to the
<strong>encapsulated message format</strong> defined below.</p>
<p>The <code class="docutils literal notranslate"><span class="pre">size</span></code> field of <code class="docutils literal notranslate"><span class="pre">Buffer</span></code> is not required to account for padding
bytes. Since this metadata can be used to communicate in-memory pointer
addresses between libraries, it is recommended to set <code class="docutils literal notranslate"><span class="pre">size</span></code> to the actual
memory size rather than the padded size.</p>
</section>
<section id="byte-order-endianness">
<h3>Byte Order (<a class="reference external" href="https://en.wikipedia.org/wiki/Endianness">Endianness</a>)<a class="headerlink" href="#byte-order-endianness" title="Permalink to this headline">#</a></h3>
<p>The Arrow format is little endian by default.</p>
<p>Serialized Schema metadata has an endianness field indicating
endianness of RecordBatches. Typically this is the endianness of the
system where the RecordBatch was generated. The main use case is
exchanging RecordBatches between systems with the same Endianness. At
first we will return an error when trying to read a Schema with an
endianness that does not match the underlying system. The reference
implementation is focused on Little Endian and provides tests for
it. Eventually we may provide automatic conversion via byte swapping.</p>
</section>
<section id="ipc-streaming-format">
<h3>IPC Streaming Format<a class="headerlink" href="#ipc-streaming-format" title="Permalink to this headline">#</a></h3>
<p>We provide a streaming protocol or “format” for record batches. It is
presented as a sequence of encapsulated messages, each of which
follows the format above. The schema comes first in the stream, and it
is the same for all of the record batches that follow. If any fields
in the schema are dictionary-encoded, one or more <code class="docutils literal notranslate"><span class="pre">DictionaryBatch</span></code>
messages will be included. <code class="docutils literal notranslate"><span class="pre">DictionaryBatch</span></code> and <code class="docutils literal notranslate"><span class="pre">RecordBatch</span></code>
messages may be interleaved, but before any dictionary key is used in
a <code class="docutils literal notranslate"><span class="pre">RecordBatch</span></code> it should be defined in a <code class="docutils literal notranslate"><span class="pre">DictionaryBatch</span></code>.</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="o">&lt;</span><span class="n">SCHEMA</span><span class="o">&gt;</span>
<span class="o">&lt;</span><span class="n">DICTIONARY</span> <span class="mi">0</span><span class="o">&gt;</span>
<span class="o">...</span>
<span class="o">&lt;</span><span class="n">DICTIONARY</span> <span class="n">k</span> <span class="o">-</span> <span class="mi">1</span><span class="o">&gt;</span>
<span class="o">&lt;</span><span class="n">RECORD</span> <span class="n">BATCH</span> <span class="mi">0</span><span class="o">&gt;</span>
<span class="o">...</span>
<span class="o">&lt;</span><span class="n">DICTIONARY</span> <span class="n">x</span> <span class="n">DELTA</span><span class="o">&gt;</span>
<span class="o">...</span>
<span class="o">&lt;</span><span class="n">DICTIONARY</span> <span class="n">y</span> <span class="n">DELTA</span><span class="o">&gt;</span>
<span class="o">...</span>
<span class="o">&lt;</span><span class="n">RECORD</span> <span class="n">BATCH</span> <span class="n">n</span> <span class="o">-</span> <span class="mi">1</span><span class="o">&gt;</span>
<span class="o">&lt;</span><span class="n">EOS</span> <span class="p">[</span><span class="n">optional</span><span class="p">]:</span> <span class="mh">0xFFFFFFFF</span> <span class="mh">0x00000000</span><span class="o">&gt;</span>
</pre></div>
</div>
<div class="admonition note">
<p class="admonition-title">Note</p>
<p>An edge-case for interleaved dictionary and record batches occurs
when the record batches contain dictionary encoded arrays that are
completely null. In this case, the dictionary for the encoded column might
appear after the first record batch.</p>
</div>
<p>When a stream reader implementation is reading a stream, after each
message, it may read the next 8 bytes to determine both if the stream
continues and the size of the message metadata that follows. Once the
message flatbuffer is read, you can then read the message body.</p>
<p>The stream writer can signal end-of-stream (EOS) either by writing 8 bytes
containing the 4-byte continuation indicator (<code class="docutils literal notranslate"><span class="pre">0xFFFFFFFF</span></code>) followed by 0
metadata length (<code class="docutils literal notranslate"><span class="pre">0x00000000</span></code>) or closing the stream interface. We
recommend the “.arrows” file extension for the streaming format although
in many cases these streams will not ever be stored as files.</p>
</section>
<section id="ipc-file-format">
<h3>IPC File Format<a class="headerlink" href="#ipc-file-format" title="Permalink to this headline">#</a></h3>
<p>We define a “file format” supporting random access that is an extension of
the stream format. The file starts and ends with a magic string <code class="docutils literal notranslate"><span class="pre">ARROW1</span></code>
(plus padding). What follows in the file is identical to the stream format.
At the end of the file, we write a <em>footer</em> containing a redundant copy of
the schema (which is a part of the streaming format) plus memory offsets and
sizes for each of the data blocks in the file. This enables random access to
any record batch in the file. See <a class="reference external" href="https://github.com/apache/arrow/blob/master/format/File.fbs">File.fbs</a> for the precise details of the
file footer.</p>
<p>Schematically we have:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="o">&lt;</span><span class="n">magic</span> <span class="n">number</span> <span class="s2">&quot;ARROW1&quot;</span><span class="o">&gt;</span>
<span class="o">&lt;</span><span class="n">empty</span> <span class="n">padding</span> <span class="nb">bytes</span> <span class="p">[</span><span class="n">to</span> <span class="mi">8</span> <span class="n">byte</span> <span class="n">boundary</span><span class="p">]</span><span class="o">&gt;</span>
<span class="o">&lt;</span><span class="n">STREAMING</span> <span class="n">FORMAT</span> <span class="k">with</span> <span class="n">EOS</span><span class="o">&gt;</span>
<span class="o">&lt;</span><span class="n">FOOTER</span><span class="o">&gt;</span>
<span class="o">&lt;</span><span class="n">FOOTER</span> <span class="n">SIZE</span><span class="p">:</span> <span class="n">int32</span><span class="o">&gt;</span>
<span class="o">&lt;</span><span class="n">magic</span> <span class="n">number</span> <span class="s2">&quot;ARROW1&quot;</span><span class="o">&gt;</span>
</pre></div>
</div>
<p>In the file format, there is no requirement that dictionary keys
should be defined in a <code class="docutils literal notranslate"><span class="pre">DictionaryBatch</span></code> before they are used in a
<code class="docutils literal notranslate"><span class="pre">RecordBatch</span></code>, as long as the keys are defined somewhere in the
file. Further more, it is invalid to have more than one <strong>non-delta</strong>
dictionary batch per dictionary ID (i.e. dictionary replacement is not
supported). Delta dictionaries are applied in the order they appear in
the file footer. We recommend the “.arrow” extension for files created with
this format.</p>
</section>
<section id="dictionary-messages">
<h3>Dictionary Messages<a class="headerlink" href="#dictionary-messages" title="Permalink to this headline">#</a></h3>
<p>Dictionaries are written in the stream and file formats as a sequence of record
batches, each having a single field. The complete semantic schema for a
sequence of record batches, therefore, consists of the schema along with all of
the dictionaries. The dictionary types are found in the schema, so it is
necessary to read the schema to first determine the dictionary types so that
the dictionaries can be properly interpreted:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="n">table</span> <span class="n">DictionaryBatch</span> <span class="p">{</span>
<span class="nb">id</span><span class="p">:</span> <span class="n">long</span><span class="p">;</span>
<span class="n">data</span><span class="p">:</span> <span class="n">RecordBatch</span><span class="p">;</span>
<span class="n">isDelta</span><span class="p">:</span> <span class="n">boolean</span> <span class="o">=</span> <span class="n">false</span><span class="p">;</span>
<span class="p">}</span>
</pre></div>
</div>
<p>The dictionary <code class="docutils literal notranslate"><span class="pre">id</span></code> in the message metadata can be referenced one or more times
in the schema, so that dictionaries can even be used for multiple fields. See
the <a class="reference internal" href="#dictionary-encoded-layout"><span class="std std-ref">Dictionary-encoded Layout</span></a> section for more about the semantics of
dictionary-encoded data.</p>
<p>The dictionary <code class="docutils literal notranslate"><span class="pre">isDelta</span></code> flag allows existing dictionaries to be
expanded for future record batch materializations. A dictionary batch
with <code class="docutils literal notranslate"><span class="pre">isDelta</span></code> set indicates that its vector should be concatenated
with those of any previous batches with the same <code class="docutils literal notranslate"><span class="pre">id</span></code>. In a stream
which encodes one column, the list of strings <code class="docutils literal notranslate"><span class="pre">[&quot;A&quot;,</span> <span class="pre">&quot;B&quot;,</span> <span class="pre">&quot;C&quot;,</span> <span class="pre">&quot;B&quot;,</span>
<span class="pre">&quot;D&quot;,</span> <span class="pre">&quot;C&quot;,</span> <span class="pre">&quot;E&quot;,</span> <span class="pre">&quot;A&quot;]</span></code>, with a delta dictionary batch could take the
form:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="o">&lt;</span><span class="n">SCHEMA</span><span class="o">&gt;</span>
<span class="o">&lt;</span><span class="n">DICTIONARY</span> <span class="mi">0</span><span class="o">&gt;</span>
<span class="p">(</span><span class="mi">0</span><span class="p">)</span> <span class="s2">&quot;A&quot;</span>
<span class="p">(</span><span class="mi">1</span><span class="p">)</span> <span class="s2">&quot;B&quot;</span>
<span class="p">(</span><span class="mi">2</span><span class="p">)</span> <span class="s2">&quot;C&quot;</span>
<span class="o">&lt;</span><span class="n">RECORD</span> <span class="n">BATCH</span> <span class="mi">0</span><span class="o">&gt;</span>
<span class="mi">0</span>
<span class="mi">1</span>
<span class="mi">2</span>
<span class="mi">1</span>
<span class="o">&lt;</span><span class="n">DICTIONARY</span> <span class="mi">0</span> <span class="n">DELTA</span><span class="o">&gt;</span>
<span class="p">(</span><span class="mi">3</span><span class="p">)</span> <span class="s2">&quot;D&quot;</span>
<span class="p">(</span><span class="mi">4</span><span class="p">)</span> <span class="s2">&quot;E&quot;</span>
<span class="o">&lt;</span><span class="n">RECORD</span> <span class="n">BATCH</span> <span class="mi">1</span><span class="o">&gt;</span>
<span class="mi">3</span>
<span class="mi">2</span>
<span class="mi">4</span>
<span class="mi">0</span>
<span class="n">EOS</span>
</pre></div>
</div>
<p>Alternatively, if <code class="docutils literal notranslate"><span class="pre">isDelta</span></code> is set to false, then the dictionary
replaces the existing dictionary for the same ID. Using the same
example as above, an alternate encoding could be:</p>
<div class="highlight-default notranslate"><div class="highlight"><pre><span></span><span class="o">&lt;</span><span class="n">SCHEMA</span><span class="o">&gt;</span>
<span class="o">&lt;</span><span class="n">DICTIONARY</span> <span class="mi">0</span><span class="o">&gt;</span>
<span class="p">(</span><span class="mi">0</span><span class="p">)</span> <span class="s2">&quot;A&quot;</span>
<span class="p">(</span><span class="mi">1</span><span class="p">)</span> <span class="s2">&quot;B&quot;</span>
<span class="p">(</span><span class="mi">2</span><span class="p">)</span> <span class="s2">&quot;C&quot;</span>
<span class="o">&lt;</span><span class="n">RECORD</span> <span class="n">BATCH</span> <span class="mi">0</span><span class="o">&gt;</span>
<span class="mi">0</span>
<span class="mi">1</span>
<span class="mi">2</span>
<span class="mi">1</span>
<span class="o">&lt;</span><span class="n">DICTIONARY</span> <span class="mi">0</span><span class="o">&gt;</span>
<span class="p">(</span><span class="mi">0</span><span class="p">)</span> <span class="s2">&quot;A&quot;</span>
<span class="p">(</span><span class="mi">1</span><span class="p">)</span> <span class="s2">&quot;C&quot;</span>
<span class="p">(</span><span class="mi">2</span><span class="p">)</span> <span class="s2">&quot;D&quot;</span>
<span class="p">(</span><span class="mi">3</span><span class="p">)</span> <span class="s2">&quot;E&quot;</span>
<span class="o">&lt;</span><span class="n">RECORD</span> <span class="n">BATCH</span> <span class="mi">1</span><span class="o">&gt;</span>
<span class="mi">2</span>
<span class="mi">1</span>
<span class="mi">3</span>
<span class="mi">0</span>
<span class="n">EOS</span>
</pre></div>
</div>
</section>
<section id="custom-application-metadata">
<h3>Custom Application Metadata<a class="headerlink" href="#custom-application-metadata" title="Permalink to this headline">#</a></h3>
<p>We provide a <code class="docutils literal notranslate"><span class="pre">custom_metadata</span></code> field at three levels to provide a
mechanism for developers to pass application-specific metadata in
Arrow protocol messages. This includes <code class="docutils literal notranslate"><span class="pre">Field</span></code>, <code class="docutils literal notranslate"><span class="pre">Schema</span></code>, and
<code class="docutils literal notranslate"><span class="pre">Message</span></code>.</p>
<p>The colon symbol <code class="docutils literal notranslate"><span class="pre">:</span></code> is to be used as a namespace separator. It can
be used multiple times in a key.</p>
<p>The <code class="docutils literal notranslate"><span class="pre">ARROW</span></code> pattern is a reserved namespace for internal Arrow use
in the <code class="docutils literal notranslate"><span class="pre">custom_metadata</span></code> fields. For example,
<code class="docutils literal notranslate"><span class="pre">ARROW:extension:name</span></code>.</p>
</section>
<section id="extension-types">
<span id="format-metadata-extension-types"></span><h3>Extension Types<a class="headerlink" href="#extension-types" title="Permalink to this headline">#</a></h3>
<p>User-defined “extension” types can be defined setting certain
<code class="docutils literal notranslate"><span class="pre">KeyValue</span></code> pairs in <code class="docutils literal notranslate"><span class="pre">custom_metadata</span></code> in the <code class="docutils literal notranslate"><span class="pre">Field</span></code> metadata
structure. These extension keys are:</p>
<ul class="simple">
<li><p><code class="docutils literal notranslate"><span class="pre">'ARROW:extension:name'</span></code> for the string name identifying the
custom data type. We recommend that you use a “namespace”-style
prefix for extension type names to minimize the possibility of
conflicts with multiple Arrow readers and writers in the same
application. For example, use <code class="docutils literal notranslate"><span class="pre">myorg.name_of_type</span></code> instead of
simply <code class="docutils literal notranslate"><span class="pre">name_of_type</span></code></p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">'ARROW:extension:metadata'</span></code> for a serialized representation
of the <code class="docutils literal notranslate"><span class="pre">ExtensionType</span></code> necessary to reconstruct the custom type</p></li>
</ul>
<p>This extension metadata can annotate any of the built-in Arrow logical
types. The intent is that an implementation that does not support an
extension type can still handle the underlying data. For example a
16-byte UUID value could be embedded in <code class="docutils literal notranslate"><span class="pre">FixedSizeBinary(16)</span></code>, and
implementations that do not have this extension type can still work
with the underlying binary values and pass along the
<code class="docutils literal notranslate"><span class="pre">custom_metadata</span></code> in subsequent Arrow protocol messages.</p>
<p>Extension types may or may not use the
<code class="docutils literal notranslate"><span class="pre">'ARROW:extension:metadata'</span></code> field. Let’s consider some example
extension types:</p>
<ul class="simple">
<li><p><code class="docutils literal notranslate"><span class="pre">uuid</span></code> represented as <code class="docutils literal notranslate"><span class="pre">FixedSizeBinary(16)</span></code> with empty metadata</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">latitude-longitude</span></code> represented as <code class="docutils literal notranslate"><span class="pre">struct&lt;latitude:</span> <span class="pre">double,</span>
<span class="pre">longitude:</span> <span class="pre">double&gt;</span></code>, and empty metadata</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">tensor</span></code> (multidimensional array) stored as <code class="docutils literal notranslate"><span class="pre">Binary</span></code> values and
having serialized metadata indicating the data type and shape of
each value. This could be JSON like <code class="docutils literal notranslate"><span class="pre">{'type':</span> <span class="pre">'int8',</span> <span class="pre">'shape':</span> <span class="pre">[4,</span>
<span class="pre">5]}</span></code> for a 4x5 cell tensor.</p></li>
<li><p><code class="docutils literal notranslate"><span class="pre">trading-time</span></code> represented as <code class="docutils literal notranslate"><span class="pre">Timestamp</span></code> with serialized
metadata indicating the market trading calendar the data corresponds
to</p></li>
</ul>
</section>
</section>
<section id="implementation-guidelines">
<h2>Implementation guidelines<a class="headerlink" href="#implementation-guidelines" title="Permalink to this headline">#</a></h2>
<p>An execution engine (or framework, or UDF executor, or storage engine,
etc) can implement only a subset of the Arrow spec and/or extend it
given the following constraints:</p>
<section id="implementing-a-subset-the-spec">
<h3>Implementing a subset the spec<a class="headerlink" href="#implementing-a-subset-the-spec" title="Permalink to this headline">#</a></h3>
<ul class="simple">
<li><p><strong>If only producing (and not consuming) arrow vectors</strong>: Any subset
of the vector spec and the corresponding metadata can be implemented.</p></li>
<li><p><strong>If consuming and producing vectors</strong>: There is a minimal subset of
vectors to be supported. Production of a subset of vectors and
their corresponding metadata is always fine. Consumption of vectors
should at least convert the unsupported input vectors to the
supported subset (for example Timestamp.millis to timestamp.micros
or int32 to int64).</p></li>
</ul>
</section>
<section id="extensibility">
<h3>Extensibility<a class="headerlink" href="#extensibility" title="Permalink to this headline">#</a></h3>
<p>An execution engine implementor can also extend their memory
representation with their own vectors internally as long as they are
never exposed. Before sending data to another system expecting Arrow
data, these custom vectors should be converted to a type that exist in
the Arrow spec.</p>
</section>
</section>
</section>
</div>
<!-- Previous / next buttons -->
<div class='prev-next-area'>
<a class='left-prev' id="prev-link" href="Versioning.html" title="previous page">
<i class="fas fa-angle-left"></i>
<div class="prev-next-info">
<p class="prev-next-subtitle">previous</p>
<p class="prev-next-title">Format Versioning and Stability</p>
</div>
</a>
<a class='right-next' id="next-link" href="Flight.html" title="next page">
<div class="prev-next-info">
<p class="prev-next-subtitle">next</p>
<p class="prev-next-title">Arrow Flight RPC</p>
</div>
<i class="fas fa-angle-right"></i>
</a>
</div>
</main>
</div>
</div>
<!-- Scripts loaded after <body> so the DOM is not blocked -->
<script src="../_static/scripts/pydata-sphinx-theme.js?digest=1999514e3f237ded88cf"></script>
<footer class="footer mt-5 mt-md-0">
<div class="container">
<div class="footer-item">
<p class="copyright">
&copy; Copyright 2016-2022 Apache Software Foundation.<br>
</p>
</div>
<div class="footer-item">
<p class="sphinx-version">
Created using <a href="http://sphinx-doc.org/">Sphinx</a> 4.5.0.<br>
</p>
</div>
</div>
</footer>
<script type="text/javascript" src="/docs/_static/versionwarning.js"></script> </body>
</html>