<!--

    Licensed to the Apache Software Foundation (ASF) under one
    or more contributor license agreements.  See the NOTICE file
    distributed with this work for additional information
    regarding copyright ownership.  The ASF licenses this file
    to you under the Apache License, Version 2.0 (the
    "License"); you may not use this file except in compliance
    with the License.  You may obtain a copy of the License at
    
        http://www.apache.org/licenses/LICENSE-2.0
    
    Unless required by applicable law or agreed to in writing,
    software distributed under the License is distributed on an
    "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
    KIND, either express or implied.  See the License for the
    specific language governing permissions and limitations
    under the License.

-->

# Database Programming

## TRIGGER

### 1. Instructions

The trigger provides a mechanism for listening to changes in time series data. With user-defined logic, tasks such as alerting and data forwarding can be conducted.

The trigger is implemented based on the reflection mechanism. Users can monitor data changes by implementing the Java interfaces. IoTDB allows users to dynamically register and drop triggers without restarting the server.

The document will help you learn to define and manage triggers.

#### Pattern for Listening

A single trigger can be used to listen for data changes in a time series that match a specific pattern. For example, a trigger can listen for the data changes of time series `root.sg.a`, or time series that match the pattern `root.sg.*`. When you register a trigger, you can specify the path pattern that the trigger listens on through an SQL statement.

#### Trigger Type

There are currently two types of triggers, and you can specify the type through an SQL statement when registering a trigger:

- Stateful triggers: The execution logic of this type of trigger may depend on data from multiple insertion statement . The framework will aggregate the data written by different nodes into the same trigger instance for calculation to retain context information. This type of trigger is usually used for sampling or statistical data aggregation for a period of time. information. Only one node in the cluster holds an instance of a stateful trigger.
- Stateless triggers: The execution logic of the trigger is only related to the current input data. The framework does not need to aggregate the data of different nodes into the same trigger instance. This type of trigger is usually used for calculation of single row data and abnormal detection. Each node in the cluster holds an instance of a stateless trigger.

#### Trigger Event

There are currently two trigger events for the trigger, and other trigger events will be expanded in the future. When you register a trigger, you can specify the trigger event through an SQL statement:

- BEFORE INSERT: Fires before the data is persisted. **Please note that currently the trigger does not support data cleaning and will not change the data to be persisted itself.**
- AFTER INSERT: Fires after the data is persisted.

### 2. How to Implement a Trigger

You need to implement the trigger by writing a Java class, where the dependency shown below is required. If you use [Maven](http://search.maven.org/), you can search for them directly from the [Maven repository](http://search.maven.org/).

#### Dependency

```xml
<dependency>
  <groupId>org.apache.iotdb</groupId>
  <artifactId>iotdb-server</artifactId>
  <version>1.0.0</version>
  <scope>provided</scope>
</dependency>
```

Note that the dependency version should be correspondent to the target server version.

#### Interface Description

To implement a trigger, you need to implement the `org.apache.iotdb.trigger.api.Trigger` class.

```java
import org.apache.iotdb.trigger.api.enums.FailureStrategy;
import org.apache.iotdb.tsfile.write.record.Tablet;

public interface Trigger {

  /**
   * This method is mainly used to validate {@link TriggerAttributes} before calling {@link
   * Trigger#onCreate(TriggerAttributes)}.
   *
   * @param attributes TriggerAttributes
   * @throws Exception e
   */
  default void validate(TriggerAttributes attributes) throws Exception {}

  /**
   * This method will be called when creating a trigger after validation.
   *
   * @param attributes TriggerAttributes
   * @throws Exception e
   */
  default void onCreate(TriggerAttributes attributes) throws Exception {}

  /**
   * This method will be called when dropping a trigger.
   *
   * @throws Exception e
   */
  default void onDrop() throws Exception {}

  /**
   * When restarting a DataNode, Triggers that have been registered will be restored and this method
   * will be called during the process of restoring.
   *
   * @throws Exception e
   */
  default void restore() throws Exception {}

  /**
   * Overrides this method to set the expected FailureStrategy, {@link FailureStrategy#OPTIMISTIC}
   * is the default strategy.
   *
   * @return {@link FailureStrategy}
   */
  default FailureStrategy getFailureStrategy() {
    return FailureStrategy.OPTIMISTIC;
  }

  /**
   * @param tablet see {@link Tablet} for detailed information of data structure. Data that is
   *     inserted will be constructed as a Tablet and you can define process logic with {@link
   *     Tablet}.
   * @return true if successfully fired
   * @throws Exception e
   */
  default boolean fire(Tablet tablet) throws Exception {
    return true;
  }
}
```

This class provides two types of programming interfaces: **Lifecycle related interfaces** and **data change listening related interfaces**. All the interfaces in this class are not required to be implemented. When the interfaces are not implemented, the trigger will not respond to the data changes. You can implement only some of these interfaces according to your needs.

Descriptions of the interfaces are as followed.

##### Lifecycle Related Interfaces

| Interface                                                    | Description                                                  |
| ------------------------------------------------------------ | ------------------------------------------------------------ |
| *default void validate(TriggerAttributes attributes) throws Exception {}* | When you creates a trigger using the `CREATE TRIGGER` statement, you can specify the parameters that the trigger needs to use, and this interface will be used to verify the correctness of the parameters。 |
| *default void onCreate(TriggerAttributes attributes) throws Exception {}* | This interface is called once when you create a trigger using the `CREATE TRIGGER` statement. During the lifetime of each trigger instance, this interface will be called only once. This interface is mainly used for the following functions: helping users to parse custom attributes in SQL statements (using `TriggerAttributes`). You can create or apply for resources, such as establishing external links, opening files, etc. |
| *default void onDrop() throws Exception {}*                  | This interface is called when you drop a trigger using the `DROP TRIGGER` statement. During the lifetime of each trigger instance, this interface will be called only once. This interface mainly has the following functions: it can perform the operation of resource release and can be used to persist the results of trigger calculations. |
| *default void restore() throws Exception {}*                 | When the DataNode is restarted, the cluster will restore the trigger instance registered on the DataNode, and this interface will be called once for stateful trigger during the process. After the DataNode where the stateful trigger instance is located goes down, the cluster will restore the trigger instance on another available DataNode, calling this interface once in the process. This interface can be used to customize recovery logic. |

##### Data Change Listening Related Interfaces

###### Listening Interface

```java
/**
   * @param tablet see {@link Tablet} for detailed information of data structure. Data that is
   *     inserted will be constructed as a Tablet and you can define process logic with {@link
   *     Tablet}.
   * @return true if successfully fired
   * @throws Exception e
   */
  default boolean fire(Tablet tablet) throws Exception {
    return true;
  }
```

When the data changes, the trigger uses the Tablet as the unit of firing operation. You can obtain the metadata and data of the corresponding sequence through Tablet, and then perform the corresponding trigger operation. If the fire process is successful, the return value should be true. If the interface returns false or throws an exception, we consider the trigger fire process as failed. When the trigger fire process fails, we will perform corresponding operations according to the listening strategy interface.

When performing an INSERT operation, for each time series in it, we will detect whether there is a trigger that listens to the path pattern, and then assemble the time series data that matches the path pattern listened by the same trigger into a new Tablet for trigger fire interface. Can be understood as:

```java
Map<PartialPath, List<Trigger>> pathToTriggerListMap => Map<Trigger, Tablet>
```

**Note that currently we do not make any guarantees about the order in which triggers fire.**

Here is an example:

Suppose there are three triggers, and the trigger event of the triggers are all BEFORE INSERT:

- Trigger1 listens on `root.sg.*`
- Trigger2 listens on `root.sg.a`
- Trigger3 listens on `root.sg.b`

Insertion statement:

```sql
insert into root.sg(time, a, b) values (1, 1, 1);
```

The time series `root.sg.a` matches Trigger1 and Trigger2, and the sequence `root.sg.b` matches Trigger1 and Trigger3, then:

- The data of `root.sg.a` and `root.sg.b` will be assembled into a new tablet1, and Trigger1.fire(tablet1) will be executed at the corresponding Trigger Event.
- The data of `root.sg.a` will be assembled into a new tablet2, and Trigger2.fire(tablet2) will be executed at the corresponding Trigger Event.
- The data of `root.sg.b` will be assembled into a new tablet3, and Trigger3.fire(tablet3) will be executed at the corresponding Trigger Event.

###### Listening Strategy Interface

When the trigger fails to fire, we will take corresponding actions according to the strategy set by the listening strategy interface. You can set `org.apache.iotdb.trigger.api.enums.FailureStrategy`. There are currently two strategies, optimistic and pessimistic:

- Optimistic strategy: The trigger that fails to fire does not affect the firing of subsequent triggers, nor does it affect the writing process, that is, we do not perform additional processing on the sequence involved in the trigger failure, only log the failure to record the failure, and finally inform user that data insertion is successful, but the trigger fire part failed.
- Pessimistic strategy: The failure trigger affects the processing of all subsequent Pipelines, that is, we believe that the firing failure of the trigger will cause all subsequent triggering processes to no longer be carried out. If the trigger event of the trigger is BEFORE INSERT, then the insertion will no longer be performed, and the insertion failure will be returned directly.

```java
 /**
   * Overrides this method to set the expected FailureStrategy, {@link FailureStrategy#OPTIMISTIC}
   * is the default strategy.
   *
   * @return {@link FailureStrategy}
   */
  default FailureStrategy getFailureStrategy() {
    return FailureStrategy.OPTIMISTIC;
  }
```

#### Example

If you use [Maven](http://search.maven.org/), you can refer to our sample project **trigger-example**.

You can find it [here](https://github.com/apache/iotdb/tree/master/example/trigger).

Here is the code from one of the sample projects:

```java
/*
 * Licensed to the Apache Software Foundation (ASF) under one
 * or more contributor license agreements.  See the NOTICE file
 * distributed with this work for additional information
 * regarding copyright ownership.  The ASF licenses this file
 * to you under the Apache License, Version 2.0 (the
 * "License"); you may not use this file except in compliance
 * with the License.  You may obtain a copy of the License at
 *
 *     http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing,
 * software distributed under the License is distributed on an
 * "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
 * KIND, either express or implied.  See the License for the
 * specific language governing permissions and limitations
 * under the License.
 */

package org.apache.iotdb.trigger;

import org.apache.iotdb.db.storageengine.trigger.sink.alertmanager.AlertManagerConfiguration;
import org.apache.iotdb.db.storageengine.trigger.sink.alertmanager.AlertManagerEvent;
import org.apache.iotdb.db.storageengine.trigger.sink.alertmanager.AlertManagerHandler;
import org.apache.iotdb.trigger.api.Trigger;
import org.apache.iotdb.trigger.api.TriggerAttributes;
import org.apache.iotdb.tsfile.file.metadata.enums.TSDataType;
import org.apache.iotdb.tsfile.write.record.Tablet;
import org.apache.iotdb.tsfile.write.schema.MeasurementSchema;

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

import java.io.IOException;
import java.util.HashMap;
import java.util.List;

public class ClusterAlertingExample implements Trigger {
  private static final Logger LOGGER = LoggerFactory.getLogger(ClusterAlertingExample.class);

  private final AlertManagerHandler alertManagerHandler = new AlertManagerHandler();

  private final AlertManagerConfiguration alertManagerConfiguration =
      new AlertManagerConfiguration("http://127.0.0.1:9093/api/v2/alerts");

  private String alertname;

  private final HashMap<String, String> labels = new HashMap<>();

  private final HashMap<String, String> annotations = new HashMap<>();

  @Override
  public void onCreate(TriggerAttributes attributes) throws Exception {
    alertname = "alert_test";

    labels.put("series", "root.ln.wf01.wt01.temperature");
    labels.put("value", "");
    labels.put("severity", "");

    annotations.put("summary", "high temperature");
    annotations.put("description", "{{.alertname}}: {{.series}} is {{.value}}");

    alertManagerHandler.open(alertManagerConfiguration);
  }

  @Override
  public void onDrop() throws IOException {
    alertManagerHandler.close();
  }

  @Override
  public boolean fire(Tablet tablet) throws Exception {
    List<MeasurementSchema> measurementSchemaList = tablet.getSchemas();
    for (int i = 0, n = measurementSchemaList.size(); i < n; i++) {
      if (measurementSchemaList.get(i).getType().equals(TSDataType.DOUBLE)) {
        // for example, we only deal with the columns of Double type
        double[] values = (double[]) tablet.values[i];
        for (double value : values) {
          if (value > 100.0) {
            LOGGER.info("trigger value > 100");
            labels.put("value", String.valueOf(value));
            labels.put("severity", "critical");
            AlertManagerEvent alertManagerEvent =
                new AlertManagerEvent(alertname, labels, annotations);
            alertManagerHandler.onEvent(alertManagerEvent);
          } else if (value > 50.0) {
            LOGGER.info("trigger value > 50");
            labels.put("value", String.valueOf(value));
            labels.put("severity", "warning");
            AlertManagerEvent alertManagerEvent =
                new AlertManagerEvent(alertname, labels, annotations);
            alertManagerHandler.onEvent(alertManagerEvent);
          }
        }
      }
    }
    return true;
  }
}
```

### 3. Trigger Management

You can create and drop a trigger through an SQL statement, and you can also query all registered triggers through an SQL statement.

**We recommend that you stop insertion while creating triggers.**

#### Create Trigger

Triggers can be registered on arbitrary path patterns. The time series registered with the trigger will be listened to by the trigger. When there is data change on the series, the corresponding fire method in the trigger will be called.

Registering a trigger can be done as follows:

1. Implement a Trigger class as described in the How to implement a Trigger chapter, assuming the class's full class name is `org.apache.iotdb.trigger.ClusterAlertingExample`
2. Package the project into a JAR package.
3. Register the trigger with an SQL statement. During the creation process, the `validate` and `onCreate` interfaces of the trigger will only be called once. For details, please refer to the chapter of How to implement a Trigger.

The complete SQL syntax is as follows:

```sql
// Create Trigger
createTrigger
    : CREATE triggerType TRIGGER triggerName=identifier triggerEventClause ON pathPattern AS className=STRING_LITERAL uriClause? triggerAttributeClause?
    ;

triggerType
    : STATELESS | STATEFUL
    ;

triggerEventClause
    : (BEFORE | AFTER) INSERT
    ;
        
uriClause
    : USING URI uri
    ;

uri
    : STRING_LITERAL
    ;
    
triggerAttributeClause
    : WITH LR_BRACKET triggerAttribute (COMMA triggerAttribute)* RR_BRACKET
    ;

triggerAttribute
    : key=attributeKey operator_eq value=attributeValue
    ;
```

Below is the explanation for the SQL syntax:

- triggerName: The trigger ID, which is globally unique and used to distinguish different triggers, is case-sensitive.
- triggerType: Trigger types are divided into two categories, STATELESS and STATEFUL.
- triggerEventClause: when the trigger fires, BEFORE INSERT and AFTER INSERT are supported now.
- pathPattern：The path pattern the trigger listens on, can contain wildcards * and **.
- className：The class name of the Trigger class.
- jarLocation: Optional. When this option is not specified, by default, we consider that the DBA has placed the JAR package required to create the trigger in the trigger_root_dir directory (configuration item, default is IOTDB_HOME/ext/trigger) of each DataNode node. When this option is specified, we will download and distribute the file resource corresponding to the URI to the trigger_root_dir/install directory of each DataNode.
- triggerAttributeClause: It is used to specify the parameters that need to be set when the trigger instance is created. This part is optional in the SQL syntax.

Here is an example SQL statement to help you understand:

```sql
CREATE STATELESS TRIGGER triggerTest
BEFORE INSERT
ON root.sg.**
AS 'org.apache.iotdb.trigger.ClusterAlertingExample'
USING URI '/jar/ClusterAlertingExample.jar'
WITH (
    "name" = "trigger",
    "limit" = "100"
)
```

The above SQL statement creates a trigger named triggerTest:

- The trigger is stateless.
- Fires before insertion.
- Listens on path pattern root.sg.**
- The implemented trigger class is named `org.apache.iotdb.trigger.ClusterAlertingExample`
- The JAR package URI is http://jar/ClusterAlertingExample.jar
- When creating the trigger instance, two parameters, name and limit, are passed in.

#### Drop Trigger

The trigger can be dropped by specifying the trigger ID. During the process of dropping the trigger, the `onDrop` interface of the trigger will be called only once.

The SQL syntax is:

```sql
// Drop Trigger
dropTrigger
  : DROP TRIGGER triggerName=identifier
;
```

Here is an example statement:

```sql
DROP TRIGGER triggerTest1
```

The above statement will drop the trigger with ID triggerTest1.

#### Show Trigger

You can query information about triggers that exist in the cluster through an SQL statement.

The SQL syntax is as follows:

```sql
SHOW TRIGGERS
```

The result set format of this statement is as follows:

| TriggerName  | Event                        | Type                 | State                                       | PathPattern | ClassName                               | NodeId                                  |
| ------------ | ---------------------------- | -------------------- | ------------------------------------------- | ----------- | --------------------------------------- | --------------------------------------- |
| triggerTest1 | BEFORE_INSERT / AFTER_INSERT | STATELESS / STATEFUL | INACTIVE / ACTIVE / DROPPING / TRANSFFERING | root.**     | org.apache.iotdb.trigger.TriggerExample | ALL(STATELESS) / DATA_NODE_ID(STATEFUL) |

#### Trigger State

During the process of creating and dropping triggers in the cluster, we maintain the states of the triggers. The following is a description of these states:

| State        | Description                                                  | Is it recommended to insert data? |
| ------------ | ------------------------------------------------------------ | --------------------------------- |
| INACTIVE     | The intermediate state of executing `CREATE TRIGGER`, the cluster has just recorded the trigger information on the ConfigNode, and the trigger has not been activated on any DataNode. | NO                                |
| ACTIVE       | Status after successful execution of `CREATE TRIGGE`, the trigger is available on all DataNodes in the cluster. | YES                               |
| DROPPING     | Intermediate state of executing `DROP TRIGGER`, the cluster is in the process of dropping the trigger. | NO                                |
| TRANSFERRING | The cluster is migrating the location of this trigger instance. | NO                                |

### 4. Notes

- The trigger takes effect from the time of registration, and does not process the existing historical data. **That is, only insertion requests that occur after the trigger is successfully registered will be listened to by the trigger. **
- The fire process of trigger is synchronous currently, so you need to ensure the efficiency of the trigger, otherwise the writing performance may be greatly affected. **You need to guarantee concurrency safety of triggers yourself**.
- Please do no register too many triggers in the cluster. Because the trigger information is fully stored in the ConfigNode, and there is a copy of the information in all DataNodes
- **It is recommended to stop writing when registering triggers**. Registering a trigger is not an atomic operation. When registering a trigger, there will be an intermediate state in which some nodes in the cluster have registered the trigger, and some nodes have not yet registered successfully. To avoid write requests on some nodes being listened to by triggers and not being listened to on some nodes, we recommend not to perform writes when registering triggers.
- When the node holding the stateful trigger instance goes down, we will try to restore the corresponding instance on another node. During the recovery process, we will call the restore interface of the trigger class once.
- The trigger JAR package has a size limit, which must be less than min(`config_node_ratis_log_appender_buffer_size_max`, 2G), where `config_node_ratis_log_appender_buffer_size_max` is a configuration item. For the specific meaning, please refer to the IOTDB configuration item description.
- **It is better not to have classes with the same full class name but different function implementations in different JAR packages.** For example, trigger1 and trigger2 correspond to resources trigger1.jar and trigger2.jar respectively. If two JAR packages contain a `org.apache.iotdb.trigger.example.AlertListener` class, when `CREATE TRIGGER` uses this class, the system will randomly load the class in one of the JAR packages, which will eventually leads the inconsistent behavior of trigger and other issues.

### 5. Configuration Parameters

| Parameter                                         | Meaning                                                      |
| ------------------------------------------------- | ------------------------------------------------------------ |
| *trigger_lib_dir*                                 | Directory to save the trigger jar package                    |
| *stateful\_trigger\_retry\_num\_when\_not\_found* | How many times will we retry to found an instance of stateful trigger on DataNodes if not found |

## CONTINUOUS QUERY (CQ)

### 1. Introduction

Continuous queries(CQ) are queries that run automatically and periodically on realtime data and store query results in other specified time series.

Users can implement sliding window streaming computing through continuous query, such as calculating the hourly average temperature of a sequence and writing it into a new sequence. Users can customize the `RESAMPLE` clause to create different sliding windows, which can achieve a certain degree of tolerance for out-of-order data.

### 2. Syntax

```sql
CREATE (CONTINUOUS QUERY | CQ) <cq_id> 
[RESAMPLE 
  [EVERY <every_interval>] 
  [BOUNDARY <execution_boundary_time>]
  [RANGE <start_time_offset>[, end_time_offset]] 
]
[TIMEOUT POLICY BLOCKED|DISCARD]
BEGIN
  SELECT CLAUSE
    INTO CLAUSE
    FROM CLAUSE
    [WHERE CLAUSE]
    [GROUP BY(<group_by_interval>[, <sliding_step>]) [, level = <level>]]
    [HAVING CLAUSE]
    [FILL {PREVIOUS | LINEAR | constant}]
    [LIMIT rowLimit OFFSET rowOffset]
    [ALIGN BY DEVICE]
END
```

> Note:
>
> 1. If there exists any time filters in WHERE CLAUSE, IoTDB will throw an error, because IoTDB will automatically generate a time range for the query each time it's executed.
> 2. GROUP BY TIME CLAUSE is different, it doesn't contain its original first display window parameter which is [start_time, end_time). It's still because IoTDB will automatically generate a time range for the query each time it's executed.
> 3. If there is no group by time clause in query, EVERY clause is required, otherwise IoTDB will throw an error.

#### Descriptions of parameters in CQ syntax

- `<cq_id>` specifies the globally unique id of CQ.
- `<every_interval>` specifies the query execution time interval. We currently support the units of ns, us, ms, s, m, h, d, w, and its value should not be lower than the minimum threshold configured by the user, which is `continuous_query_min_every_interval`. It's an optional parameter, default value is set to `group_by_interval` in group by clause.
- `<start_time_offset>` specifies the start time of each query execution as `now()-<start_time_offset>`. We currently support the units of ns, us, ms, s, m, h, d, w.It's an optional parameter, default value is set to `every_interval` in resample clause.
- `<end_time_offset>` specifies the end time of each query execution as `now()-<end_time_offset>`. We currently support the units of ns, us, ms, s, m, h, d, w.It's an optional parameter, default value is set to `0`.
- `<execution_boundary_time>` is a date that represents the execution time of a certain cq task.
    - `<execution_boundary_time>` can be earlier than, equals to, later than **current time**.
    - This parameter is optional. If not specified, it is equal to `BOUNDARY 0`。
    - **The start time of the first time window** is `<execution_boundary_time> - <start_time_offset>`.
    - **The end time of the first time window** is `<execution_boundary_time> - <end_time_offset>`.
    - The **time range** of the `i (1 <= i)th` window is `[<execution_boundary_time> - <start_time_offset> + (i - 1) * <every_interval>, <execution_boundary_time> - <end_time_offset> + (i - 1) * <every_interval>)`.
    - If the **current time** is earlier than or equal to `execution_boundary_time`, then the first execution moment of the continuous query is `execution_boundary_time`.
    - If the **current time** is later than `execution_boundary_time`, then the first execution moment of the continuous query is the first `execution_boundary_time + i * <every_interval>` that is later than or equal to the current time .

> - `<every_interval>`，`<start_time_offset>` and `<group_by_interval>` should all be greater than `0`.
> - The value of `<group_by_interval>` should be less than or equal to the value of `<start_time_offset>`, otherwise the system will throw an error.
> - Users should specify the appropriate `<start_time_offset>` and `<every_interval>` according to actual needs.
>     - If `<start_time_offset>` is greater than `<every_interval>`,  there will be partial data overlap in each query window.
>     - If `<start_time_offset>` is less than `<every_interval>`, there may be uncovered data between each query window.
> - `start_time_offset` should be larger than `end_time_offset`, otherwise the system will throw an error.

##### `<start_time_offset>` == `<every_interval>`

![1](https://alioss.timecho.com/docs/img/UserGuide/Process-Data/Continuous-Query/pic1.png?raw=true)

##### `<start_time_offset>` > `<every_interval>`

![2](https://alioss.timecho.com/docs/img/UserGuide/Process-Data/Continuous-Query/pic2.png?raw=true)

##### `<start_time_offset>` < `<every_interval>`

![3](https://alioss.timecho.com/docs/img/UserGuide/Process-Data/Continuous-Query/pic3.png?raw=true)

##### `<every_interval>`  is not zero

![4](https://alioss.timecho.com/docs/img/UserGuide/Process-Data/Continuous-Query/pic4.png?raw=true)


- `TIMEOUT POLICY` specify how we deal with the cq task whose previous time interval execution is not finished while the next execution time has reached. The default value is `BLOCKED`.
    - `BLOCKED` means that we will block and wait to do the current cq execution task until the previous time interval cq task finishes. If using `BLOCKED` policy, all the time intervals will be executed, but it may be behind the latest time interval.
    - `DISCARD` means that we just discard the current cq execution task and wait for the next execution time and do the next time interval cq task. If using `DISCARD` policy, some time intervals won't be executed when the execution time of one cq task is longer than the `<every_interval>`. However, once a cq task is executed, it will use the latest time interval, so it can catch up at the sacrifice of some time intervals being discarded.


### 3. Examples of CQ

The examples below use the following sample data. It's a real time data stream and we can assume that the data arrives on time.

````
+-----------------------------+-----------------------------+-----------------------------+-----------------------------+-----------------------------+
|                         Time|root.ln.wf02.wt02.temperature|root.ln.wf02.wt01.temperature|root.ln.wf01.wt02.temperature|root.ln.wf01.wt01.temperature|
+-----------------------------+-----------------------------+-----------------------------+-----------------------------+-----------------------------+
|2021-05-11T22:18:14.598+08:00|                        121.0|                         72.0|                        183.0|                        115.0|
|2021-05-11T22:18:19.941+08:00|                          0.0|                         68.0|                         68.0|                        103.0|
|2021-05-11T22:18:24.949+08:00|                        122.0|                         45.0|                         11.0|                         14.0|
|2021-05-11T22:18:29.967+08:00|                         47.0|                         14.0|                         59.0|                        181.0|
|2021-05-11T22:18:34.979+08:00|                        182.0|                        113.0|                         29.0|                        180.0|
|2021-05-11T22:18:39.990+08:00|                         42.0|                         11.0|                         52.0|                         19.0|
|2021-05-11T22:18:44.995+08:00|                         78.0|                         38.0|                        123.0|                         52.0|
|2021-05-11T22:18:49.999+08:00|                        137.0|                        172.0|                        135.0|                        193.0|
|2021-05-11T22:18:55.003+08:00|                         16.0|                        124.0|                        183.0|                         18.0|
+-----------------------------+-----------------------------+-----------------------------+-----------------------------+-----------------------------+
````

#### Configuring execution intervals

Use an `EVERY` interval in the `RESAMPLE` clause to specify the CQ’s execution interval, if not specific, default value is equal to `group_by_interval`.

```sql
CREATE CONTINUOUS QUERY cq1
RESAMPLE EVERY 20s
BEGIN
SELECT max_value(temperature)
  INTO root.ln.wf02.wt02(temperature_max), root.ln.wf02.wt01(temperature_max), root.ln.wf01.wt02(temperature_max), root.ln.wf01.wt01(temperature_max)
  FROM root.ln.*.*
  GROUP BY(10s)
END
```

`cq1` calculates the 10-second average of `temperature` sensor under the `root.ln` prefix path and stores the results in the `temperature_max` sensor using the same prefix path as the corresponding sensor.

`cq1` executes at 20-second intervals, the same interval as the `EVERY` interval. Every 20 seconds, `cq1` runs a single query that covers the time range for the current time bucket, that is, the 20-second time bucket that intersects with `now()`.

Supposing that the current time is `2021-05-11T22:18:40.000+08:00`, we can see annotated log output about `cq1` running at DataNode if you set log level to DEBUG:

````
At **2021-05-11T22:18:40.000+08:00**, `cq1` executes a query within the time range `[2021-05-11T22:18:20, 2021-05-11T22:18:40)`.
`cq1` generate 2 lines:
>
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|                         Time|root.ln.wf02.wt02.temperature_max|root.ln.wf02.wt01.temperature_max|root.ln.wf01.wt02.temperature_max|root.ln.wf01.wt01.temperature_max|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|2021-05-11T22:18:20.000+08:00|                            122.0|                             45.0|                             59.0|                            181.0|
|2021-05-11T22:18:30.000+08:00|                            182.0|                            113.0|                             52.0|                            180.0|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
>
At **2021-05-11T22:19:00.000+08:00**, `cq1` executes a query within the time range `[2021-05-11T22:18:40, 2021-05-11T22:19:00)`.
`cq1` generate 2 lines:
>
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|                         Time|root.ln.wf02.wt02.temperature_max|root.ln.wf02.wt01.temperature_max|root.ln.wf01.wt02.temperature_max|root.ln.wf01.wt01.temperature_max|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|2021-05-11T22:18:40.000+08:00|                            137.0|                            172.0|                            135.0|                            193.0|
|2021-05-11T22:18:50.000+08:00|                             16.0|                            124.0|                            183.0|                             18.0|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
>
````

`cq1` won't deal with data that is before the current time window which is `2021-05-11T22:18:20.000+08:00`, so here are the results:

````
> SELECT temperature_max from root.ln.*.*;
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|                         Time|root.ln.wf02.wt02.temperature_max|root.ln.wf02.wt01.temperature_max|root.ln.wf01.wt02.temperature_max|root.ln.wf01.wt01.temperature_max|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|2021-05-11T22:18:20.000+08:00|                            122.0|                             45.0|                             59.0|                            181.0|
|2021-05-11T22:18:30.000+08:00|                            182.0|                            113.0|                             52.0|                            180.0|
|2021-05-11T22:18:40.000+08:00|                            137.0|                            172.0|                            135.0|                            193.0|
|2021-05-11T22:18:50.000+08:00|                             16.0|                            124.0|                            183.0|                             18.0|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
````

#### Configuring time range for resampling

Use `start_time_offset` in the `RANGE` clause to specify the start time of the CQ’s time range, if not specific, default value is equal to `EVERY` interval.

```sql
CREATE CONTINUOUS QUERY cq2
RESAMPLE RANGE 40s
BEGIN
  SELECT max_value(temperature)
  INTO root.ln.wf02.wt02(temperature_max), root.ln.wf02.wt01(temperature_max), root.ln.wf01.wt02(temperature_max), root.ln.wf01.wt01(temperature_max)
  FROM root.ln.*.*
  GROUP BY(10s)
END
```

`cq2` calculates the 10-second average of `temperature` sensor under the `root.ln` prefix path and stores the results in the `temperature_max` sensor using the same prefix path as the corresponding sensor.

`cq2` executes at 10-second intervals, the same interval as the `group_by_interval`. Every 10 seconds, `cq2` runs a single query that covers the time range between `now()` minus the `start_time_offset` and `now()` , that is, the time range between 40 seconds prior to `now()` and `now()`.

Supposing that the current time is `2021-05-11T22:18:40.000+08:00`, we can see annotated log output about `cq2` running at DataNode if you set log level to DEBUG:

````
At **2021-05-11T22:18:40.000+08:00**, `cq2` executes a query within the time range `[2021-05-11T22:18:00, 2021-05-11T22:18:40)`.
`cq2` generate 4 lines:
>
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|                         Time|root.ln.wf02.wt02.temperature_max|root.ln.wf02.wt01.temperature_max|root.ln.wf01.wt02.temperature_max|root.ln.wf01.wt01.temperature_max|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|2021-05-11T22:18:00.000+08:00|                             NULL|                             NULL|                             NULL|                             NULL|
|2021-05-11T22:18:10.000+08:00|                            121.0|                             72.0|                            183.0|                            115.0|
|2021-05-11T22:18:20.000+08:00|                            122.0|                             45.0|                             59.0|                            181.0|
|2021-05-11T22:18:30.000+08:00|                            182.0|                            113.0|                             52.0|                            180.0|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
>
At **2021-05-11T22:18:50.000+08:00**, `cq2` executes a query within the time range `[2021-05-11T22:18:10, 2021-05-11T22:18:50)`.
`cq2` generate 4 lines:
>
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|                         Time|root.ln.wf02.wt02.temperature_max|root.ln.wf02.wt01.temperature_max|root.ln.wf01.wt02.temperature_max|root.ln.wf01.wt01.temperature_max|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|2021-05-11T22:18:10.000+08:00|                            121.0|                             72.0|                            183.0|                            115.0|
|2021-05-11T22:18:20.000+08:00|                            122.0|                             45.0|                             59.0|                            181.0|
|2021-05-11T22:18:30.000+08:00|                            182.0|                            113.0|                             52.0|                            180.0|
|2021-05-11T22:18:40.000+08:00|                            137.0|                            172.0|                            135.0|                            193.0|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
>
At **2021-05-11T22:19:00.000+08:00**, `cq2` executes a query within the time range `[2021-05-11T22:18:20, 2021-05-11T22:19:00)`.
`cq2` generate 4 lines:
>
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|                         Time|root.ln.wf02.wt02.temperature_max|root.ln.wf02.wt01.temperature_max|root.ln.wf01.wt02.temperature_max|root.ln.wf01.wt01.temperature_max|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|2021-05-11T22:18:20.000+08:00|                            122.0|                             45.0|                             59.0|                            181.0|
|2021-05-11T22:18:30.000+08:00|                            182.0|                            113.0|                             52.0|                            180.0|
|2021-05-11T22:18:40.000+08:00|                            137.0|                            172.0|                            135.0|                            193.0|
|2021-05-11T22:18:50.000+08:00|                             16.0|                            124.0|                            183.0|                             18.0|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
>
````

`cq2` won't write lines that are all null. Notice `cq2` will also calculate the results for some time interval many times. Here are the results:

````
> SELECT temperature_max from root.ln.*.*;
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|                         Time|root.ln.wf02.wt02.temperature_max|root.ln.wf02.wt01.temperature_max|root.ln.wf01.wt02.temperature_max|root.ln.wf01.wt01.temperature_max|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|2021-05-11T22:18:10.000+08:00|                            121.0|                             72.0|                            183.0|                            115.0|
|2021-05-11T22:18:20.000+08:00|                            122.0|                             45.0|                             59.0|                            181.0|
|2021-05-11T22:18:30.000+08:00|                            182.0|                            113.0|                             52.0|                            180.0|
|2021-05-11T22:18:40.000+08:00|                            137.0|                            172.0|                            135.0|                            193.0|
|2021-05-11T22:18:50.000+08:00|                             16.0|                            124.0|                            183.0|                             18.0|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
````

#### Configuring execution intervals and CQ time ranges

Use an `EVERY` interval and `RANGE` interval in the `RESAMPLE` clause to specify the CQ’s execution interval and the length of the CQ’s time range. And use `fill()` to change the value reported for time intervals with no data.

```sql
CREATE CONTINUOUS QUERY cq3
RESAMPLE EVERY 20s RANGE 40s
BEGIN
  SELECT max_value(temperature)
  INTO root.ln.wf02.wt02(temperature_max), root.ln.wf02.wt01(temperature_max), root.ln.wf01.wt02(temperature_max), root.ln.wf01.wt01(temperature_max)
  FROM root.ln.*.*
  GROUP BY(10s)
  FILL(100.0)
END
```

`cq3` calculates the 10-second average of `temperature` sensor under the `root.ln` prefix path and stores the results in the `temperature_max` sensor using the same prefix path as the corresponding sensor. Where possible, it writes the value `100.0` for time intervals with no results.

`cq3` executes at 20-second intervals, the same interval as the `EVERY` interval. Every 20 seconds, `cq3` runs a single query that covers the time range between `now()` minus the `start_time_offset` and `now()`, that is, the time range between 40 seconds prior to `now()` and `now()`.

Supposing that the current time is `2021-05-11T22:18:40.000+08:00`, we can see annotated log output about `cq3` running at DataNode if you set log level to DEBUG:

````
At **2021-05-11T22:18:40.000+08:00**, `cq3` executes a query within the time range `[2021-05-11T22:18:00, 2021-05-11T22:18:40)`.
`cq3` generate 4 lines:
>
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|                         Time|root.ln.wf02.wt02.temperature_max|root.ln.wf02.wt01.temperature_max|root.ln.wf01.wt02.temperature_max|root.ln.wf01.wt01.temperature_max|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|2021-05-11T22:18:00.000+08:00|                            100.0|                            100.0|                            100.0|                            100.0|
|2021-05-11T22:18:10.000+08:00|                            121.0|                             72.0|                            183.0|                            115.0|
|2021-05-11T22:18:20.000+08:00|                            122.0|                             45.0|                             59.0|                            181.0|
|2021-05-11T22:18:30.000+08:00|                            182.0|                            113.0|                             52.0|                            180.0|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
>
At **2021-05-11T22:19:00.000+08:00**, `cq3` executes a query within the time range `[2021-05-11T22:18:20, 2021-05-11T22:19:00)`.
`cq3` generate 4 lines:
>
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|                         Time|root.ln.wf02.wt02.temperature_max|root.ln.wf02.wt01.temperature_max|root.ln.wf01.wt02.temperature_max|root.ln.wf01.wt01.temperature_max|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|2021-05-11T22:18:20.000+08:00|                            122.0|                             45.0|                             59.0|                            181.0|
|2021-05-11T22:18:30.000+08:00|                            182.0|                            113.0|                             52.0|                            180.0|
|2021-05-11T22:18:40.000+08:00|                            137.0|                            172.0|                            135.0|                            193.0|
|2021-05-11T22:18:50.000+08:00|                             16.0|                            124.0|                            183.0|                             18.0|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
>
````

Notice that `cq3` will calculate the results for some time interval many times, so here are the results:

````
> SELECT temperature_max from root.ln.*.*;
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|                         Time|root.ln.wf02.wt02.temperature_max|root.ln.wf02.wt01.temperature_max|root.ln.wf01.wt02.temperature_max|root.ln.wf01.wt01.temperature_max|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|2021-05-11T22:18:00.000+08:00|                            100.0|                            100.0|                            100.0|                            100.0|
|2021-05-11T22:18:10.000+08:00|                            121.0|                             72.0|                            183.0|                            115.0|
|2021-05-11T22:18:20.000+08:00|                            122.0|                             45.0|                             59.0|                            181.0|
|2021-05-11T22:18:30.000+08:00|                            182.0|                            113.0|                             52.0|                            180.0|
|2021-05-11T22:18:40.000+08:00|                            137.0|                            172.0|                            135.0|                            193.0|
|2021-05-11T22:18:50.000+08:00|                             16.0|                            124.0|                            183.0|                             18.0|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
````

#### Configuring end_time_offset for CQ time range

Use an `EVERY` interval and `RANGE` interval in the RESAMPLE clause to specify the CQ’s execution interval and the length of the CQ’s time range. And use `fill()` to change the value reported for time intervals with no data.

```sql
CREATE CONTINUOUS QUERY cq4
RESAMPLE EVERY 20s RANGE 40s, 20s
BEGIN
  SELECT max_value(temperature)
  INTO root.ln.wf02.wt02(temperature_max), root.ln.wf02.wt01(temperature_max), root.ln.wf01.wt02(temperature_max), root.ln.wf01.wt01(temperature_max)
  FROM root.ln.*.*
  GROUP BY(10s)
  FILL(100.0)
END
```

`cq4` calculates the 10-second average of `temperature` sensor under the `root.ln` prefix path and stores the results in the `temperature_max` sensor using the same prefix path as the corresponding sensor. Where possible, it writes the value `100.0` for time intervals with no results.

`cq4` executes at 20-second intervals, the same interval as the `EVERY` interval. Every 20 seconds, `cq4` runs a single query that covers the time range between `now()` minus the `start_time_offset` and `now()` minus the `end_time_offset`, that is, the time range between  40 seconds prior to `now()` and 20 seconds prior to `now()`.

Supposing that the current time is `2021-05-11T22:18:40.000+08:00`, we can see annotated log output about `cq4` running at DataNode if you set log level to DEBUG:

````
At **2021-05-11T22:18:40.000+08:00**, `cq4` executes a query within the time range `[2021-05-11T22:18:00, 2021-05-11T22:18:20)`.
`cq4` generate 2 lines:
>
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|                         Time|root.ln.wf02.wt02.temperature_max|root.ln.wf02.wt01.temperature_max|root.ln.wf01.wt02.temperature_max|root.ln.wf01.wt01.temperature_max|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|2021-05-11T22:18:00.000+08:00|                            100.0|                            100.0|                            100.0|                            100.0|
|2021-05-11T22:18:10.000+08:00|                            121.0|                             72.0|                            183.0|                            115.0|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
>
At **2021-05-11T22:19:00.000+08:00**, `cq4` executes a query within the time range `[2021-05-11T22:18:20, 2021-05-11T22:18:40)`.
`cq4` generate 2 lines:
>
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|                         Time|root.ln.wf02.wt02.temperature_max|root.ln.wf02.wt01.temperature_max|root.ln.wf01.wt02.temperature_max|root.ln.wf01.wt01.temperature_max|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|2021-05-11T22:18:20.000+08:00|                            122.0|                             45.0|                             59.0|                            181.0|
|2021-05-11T22:18:30.000+08:00|                            182.0|                            113.0|                             52.0|                            180.0|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
>
````

Notice that `cq4` will calculate the results for all time intervals only once after a delay of 20 seconds, so here are the results:

````
> SELECT temperature_max from root.ln.*.*;
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|                         Time|root.ln.wf02.wt02.temperature_max|root.ln.wf02.wt01.temperature_max|root.ln.wf01.wt02.temperature_max|root.ln.wf01.wt01.temperature_max|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
|2021-05-11T22:18:00.000+08:00|                            100.0|                            100.0|                            100.0|                            100.0|
|2021-05-11T22:18:10.000+08:00|                            121.0|                             72.0|                            183.0|                            115.0|
|2021-05-11T22:18:20.000+08:00|                            122.0|                             45.0|                             59.0|                            181.0|
|2021-05-11T22:18:30.000+08:00|                            182.0|                            113.0|                             52.0|                            180.0|
+-----------------------------+---------------------------------+---------------------------------+---------------------------------+---------------------------------+
````

#### CQ without group by clause

Use an `EVERY` interval in the `RESAMPLE` clause to specify the CQ’s execution interval and the length of the CQ’s time range.

```sql
CREATE CONTINUOUS QUERY cq5
RESAMPLE EVERY 20s
BEGIN
  SELECT temperature + 1
  INTO root.precalculated_sg.::(temperature)
  FROM root.ln.*.*
  align by device
END
```

`cq5` calculates the `temperature + 1` under the `root.ln` prefix path and stores the results in the `root.precalculated_sg` database. Sensors use the same prefix path as the corresponding sensor.

`cq5` executes at 20-second intervals, the same interval as the `EVERY` interval. Every 20 seconds, `cq5` runs a single query that covers the time range for the current time bucket, that is, the 20-second time bucket that intersects with `now()`.

Supposing that the current time is `2021-05-11T22:18:40.000+08:00`, we can see annotated log output about `cq5` running at DataNode if you set log level to DEBUG:

````
At **2021-05-11T22:18:40.000+08:00**, `cq5` executes a query within the time range `[2021-05-11T22:18:20, 2021-05-11T22:18:40)`.
`cq5` generate 16 lines:
>
+-----------------------------+-------------------------------+-----------+
|                         Time|                         Device|temperature|
+-----------------------------+-------------------------------+-----------+
|2021-05-11T22:18:24.949+08:00|root.precalculated_sg.wf02.wt02|      123.0| 
|2021-05-11T22:18:29.967+08:00|root.precalculated_sg.wf02.wt02|       48.0|
|2021-05-11T22:18:34.979+08:00|root.precalculated_sg.wf02.wt02|      183.0|
|2021-05-11T22:18:39.990+08:00|root.precalculated_sg.wf02.wt02|       45.0|
|2021-05-11T22:18:24.949+08:00|root.precalculated_sg.wf02.wt01|       46.0| 
|2021-05-11T22:18:29.967+08:00|root.precalculated_sg.wf02.wt01|       15.0|
|2021-05-11T22:18:34.979+08:00|root.precalculated_sg.wf02.wt01|      114.0|
|2021-05-11T22:18:39.990+08:00|root.precalculated_sg.wf02.wt01|       12.0|
|2021-05-11T22:18:24.949+08:00|root.precalculated_sg.wf01.wt02|       12.0| 
|2021-05-11T22:18:29.967+08:00|root.precalculated_sg.wf01.wt02|       60.0|
|2021-05-11T22:18:34.979+08:00|root.precalculated_sg.wf01.wt02|       30.0|
|2021-05-11T22:18:39.990+08:00|root.precalculated_sg.wf01.wt02|       53.0|
|2021-05-11T22:18:24.949+08:00|root.precalculated_sg.wf01.wt01|       15.0| 
|2021-05-11T22:18:29.967+08:00|root.precalculated_sg.wf01.wt01|      182.0|
|2021-05-11T22:18:34.979+08:00|root.precalculated_sg.wf01.wt01|      181.0|
|2021-05-11T22:18:39.990+08:00|root.precalculated_sg.wf01.wt01|       20.0|
+-----------------------------+-------------------------------+-----------+
>
At **2021-05-11T22:19:00.000+08:00**, `cq5` executes a query within the time range `[2021-05-11T22:18:40, 2021-05-11T22:19:00)`.
`cq5` generate 12 lines:
>
+-----------------------------+-------------------------------+-----------+
|                         Time|                         Device|temperature|
+-----------------------------+-------------------------------+-----------+
|2021-05-11T22:18:44.995+08:00|root.precalculated_sg.wf02.wt02|       79.0| 
|2021-05-11T22:18:49.999+08:00|root.precalculated_sg.wf02.wt02|      138.0|
|2021-05-11T22:18:55.003+08:00|root.precalculated_sg.wf02.wt02|       17.0|
|2021-05-11T22:18:44.995+08:00|root.precalculated_sg.wf02.wt01|       39.0| 
|2021-05-11T22:18:49.999+08:00|root.precalculated_sg.wf02.wt01|      173.0|
|2021-05-11T22:18:55.003+08:00|root.precalculated_sg.wf02.wt01|      125.0|
|2021-05-11T22:18:44.995+08:00|root.precalculated_sg.wf01.wt02|      124.0| 
|2021-05-11T22:18:49.999+08:00|root.precalculated_sg.wf01.wt02|      136.0|
|2021-05-11T22:18:55.003+08:00|root.precalculated_sg.wf01.wt02|      184.0|
|2021-05-11T22:18:44.995+08:00|root.precalculated_sg.wf01.wt01|       53.0| 
|2021-05-11T22:18:49.999+08:00|root.precalculated_sg.wf01.wt01|      194.0|
|2021-05-11T22:18:55.003+08:00|root.precalculated_sg.wf01.wt01|       19.0|
+-----------------------------+-------------------------------+-----------+
>
````

`cq5` won't deal with data that is before the current time window which is `2021-05-11T22:18:20.000+08:00`, so here are the results:

````
> SELECT temperature from root.precalculated_sg.*.* align by device;
+-----------------------------+-------------------------------+-----------+
|                         Time|                         Device|temperature|
+-----------------------------+-------------------------------+-----------+
|2021-05-11T22:18:24.949+08:00|root.precalculated_sg.wf02.wt02|      123.0| 
|2021-05-11T22:18:29.967+08:00|root.precalculated_sg.wf02.wt02|       48.0|
|2021-05-11T22:18:34.979+08:00|root.precalculated_sg.wf02.wt02|      183.0|
|2021-05-11T22:18:39.990+08:00|root.precalculated_sg.wf02.wt02|       45.0|
|2021-05-11T22:18:44.995+08:00|root.precalculated_sg.wf02.wt02|       79.0| 
|2021-05-11T22:18:49.999+08:00|root.precalculated_sg.wf02.wt02|      138.0|
|2021-05-11T22:18:55.003+08:00|root.precalculated_sg.wf02.wt02|       17.0|
|2021-05-11T22:18:24.949+08:00|root.precalculated_sg.wf02.wt01|       46.0| 
|2021-05-11T22:18:29.967+08:00|root.precalculated_sg.wf02.wt01|       15.0|
|2021-05-11T22:18:34.979+08:00|root.precalculated_sg.wf02.wt01|      114.0|
|2021-05-11T22:18:39.990+08:00|root.precalculated_sg.wf02.wt01|       12.0|
|2021-05-11T22:18:44.995+08:00|root.precalculated_sg.wf02.wt01|       39.0| 
|2021-05-11T22:18:49.999+08:00|root.precalculated_sg.wf02.wt01|      173.0|
|2021-05-11T22:18:55.003+08:00|root.precalculated_sg.wf02.wt01|      125.0|
|2021-05-11T22:18:24.949+08:00|root.precalculated_sg.wf01.wt02|       12.0| 
|2021-05-11T22:18:29.967+08:00|root.precalculated_sg.wf01.wt02|       60.0|
|2021-05-11T22:18:34.979+08:00|root.precalculated_sg.wf01.wt02|       30.0|
|2021-05-11T22:18:39.990+08:00|root.precalculated_sg.wf01.wt02|       53.0|
|2021-05-11T22:18:44.995+08:00|root.precalculated_sg.wf01.wt02|      124.0| 
|2021-05-11T22:18:49.999+08:00|root.precalculated_sg.wf01.wt02|      136.0|
|2021-05-11T22:18:55.003+08:00|root.precalculated_sg.wf01.wt02|      184.0|
|2021-05-11T22:18:24.949+08:00|root.precalculated_sg.wf01.wt01|       15.0| 
|2021-05-11T22:18:29.967+08:00|root.precalculated_sg.wf01.wt01|      182.0|
|2021-05-11T22:18:34.979+08:00|root.precalculated_sg.wf01.wt01|      181.0|
|2021-05-11T22:18:39.990+08:00|root.precalculated_sg.wf01.wt01|       20.0|
|2021-05-11T22:18:44.995+08:00|root.precalculated_sg.wf01.wt01|       53.0| 
|2021-05-11T22:18:49.999+08:00|root.precalculated_sg.wf01.wt01|      194.0|
|2021-05-11T22:18:55.003+08:00|root.precalculated_sg.wf01.wt01|       19.0|
+-----------------------------+-------------------------------+-----------+
````

### 4. CQ Management

#### Listing continuous queries

List every CQ on the IoTDB Cluster with:

```sql
SHOW (CONTINUOUS QUERIES | CQS) 
```

`SHOW (CONTINUOUS QUERIES | CQS)` order results by `cq_id`.

##### Examples

```sql
SHOW CONTINUOUS QUERIES;
```

we will get:

| cq_id       | query                                                        | state  |
| :---------- | ------------------------------------------------------------ | ------ |
| s1_count_cq | CREATE CQ s1_count_cq<br/>BEGIN<br/>SELECT count(s1)<br/>INTO root.sg_count.d.count_s1<br/>FROM root.sg.d<br/>GROUP BY(30m)<br/>END | active |


#### Dropping continuous queries

Drop a CQ with a specific `cq_id`:

```sql
DROP (CONTINUOUS QUERY | CQ) <cq_id>
```

DROP CQ returns an empty result.

##### Examples

Drop the CQ named `s1_count_cq`:

```sql
DROP CONTINUOUS QUERY s1_count_cq;
```

#### Altering continuous queries

CQs can't be altered once they're created. To change a CQ, you must `DROP` and re`CREATE` it with the updated settings.


### 5. CQ Use Cases

#### Downsampling and Data Retention

Use CQs with `TTL` set on database in IoTDB to mitigate storage concerns. Combine CQs and `TTL` to automatically downsample high precision data to a lower precision and remove the dispensable, high precision data from the database.

#### Recalculating expensive queries

Shorten query runtimes by pre-calculating expensive queries with CQs. Use a CQ to automatically downsample commonly-queried, high precision data to a lower precision. Queries on lower precision data require fewer resources and return faster.

> Pre-calculate queries for your preferred graphing tool to accelerate the population of graphs and dashboards.

#### Substituting for sub-query

IoTDB does not support sub queries. We can get the same functionality by creating a CQ as a sub query and store its result into other time series and then querying from those time series again will be like doing nested sub query.

##### Example

IoTDB does not accept the following query with a nested sub query. The query calculates the average number of non-null values of `s1` at 30 minute intervals:

```sql
SELECT avg(count_s1) from (select count(s1) as count_s1 from root.sg.d group by([0, now()), 30m));
```

To get the same results:

**1. Create a CQ**

This step performs the nested sub query in from clause of the query above. The following CQ automatically calculates the number of non-null values of `s1` at 30 minute intervals and writes those counts into the new `root.sg_count.d.count_s1` time series.

```sql
CREATE CQ s1_count_cq 
BEGIN 
    SELECT count(s1)  
        INTO root.sg_count.d(count_s1)
        FROM root.sg.d
        GROUP BY(30m)
END
```

**2. Query the CQ results**

Next step performs the avg([...]) part of the outer query above.

Query the data in the time series `root.sg_count.d.count_s1` to calculate the average of it:

```sql
SELECT avg(count_s1) from root.sg_count.d;
```


### 6. System Parameter Configuration

| Name                                        | Description                                                  | Data Type | Default Value |
| :------------------------------------------ | ------------------------------------------------------------ | --------- | ------------- |
| `continuous_query_submit_thread`            | The number of threads in the scheduled thread pool that submit continuous query tasks periodically | int32     | 2             |
| `continuous_query_min_every_interval_in_ms` | The minimum value of the continuous query execution time interval | duration  | 1000          |

## USER-DEFINED FUNCTION (UDF)

IoTDB provides a variety of built-in functions to meet your computing needs, and you can also create user defined functions to meet more computing needs. 

This document describes how to write, register and use a UDF.


### UDF Types

In IoTDB, you can expand two types of UDF:

| UDF Class                                           | Description                                                  |
| --------------------------------------------------- | ------------------------------------------------------------ |
| UDTF（User Defined Timeseries Generating Function） | This type of function can take **multiple** time series as input, and output **one** time series, which can have any number of data points. |
| UDAF（User Defined Aggregation Function）           | Custom Aggregation Functions. This type of function can take one time series as input, and output **one** aggregated data point for each group based on the GROUP BY type. |

### UDF Development Dependencies

If you use [Maven](http://search.maven.org/), you can search for the development dependencies listed below from the [Maven repository](http://search.maven.org/) . Please note that you must select the same dependency version as the target IoTDB server version for development.

``` xml
<dependency>
  <groupId>org.apache.iotdb</groupId>
  <artifactId>udf-api</artifactId>
  <version>1.0.0</version>
  <scope>provided</scope>
</dependency>
```

### UDTF（User Defined Timeseries Generating Function）

To write a UDTF,  you need to inherit the `org.apache.iotdb.udf.api.UDTF` class, and at least implement the `beforeStart` method and a `transform` method.

The following table shows all the interfaces available for user implementation. 

| Interface definition                                         | Description                                                  | Required to Implement                                 |
| :----------------------------------------------------------- | :----------------------------------------------------------- | ----------------------------------------------------- |
| `void validate(UDFParameterValidator validator) throws Exception` | This method is mainly used to validate `UDFParameters` and it is executed before `beforeStart(UDFParameters, UDTFConfigurations)` is called. | Optional                                              |
| `void beforeStart(UDFParameters parameters, UDTFConfigurations configurations) throws Exception` | The initialization method to call the user-defined initialization behavior before a UDTF processes the input data. Every time a user executes a UDTF query, the framework will construct a new UDF instance, and `beforeStart` will be called. | Required                                              |
| `void transform(Row row, PointCollector collector) throws Exception` | This method is called by the framework. This data processing method will be called when you choose to use the `RowByRowAccessStrategy` strategy (set in `beforeStart`) to consume raw data. Input data is passed in by `Row`, and the transformation result should be output by `PointCollector`. You need to call the data collection method provided by `collector`  to determine the output data. | Required to implement at least one `transform` method |
| `void transform(RowWindow rowWindow, PointCollector collector) throws Exception` | This method is called by the framework. This data processing method will be called when you choose to use the `SlidingSizeWindowAccessStrategy` or `SlidingTimeWindowAccessStrategy` strategy (set in `beforeStart`) to consume raw data. Input data is passed in by `RowWindow`, and the transformation result should be output by `PointCollector`. You need to call the data collection method provided by `collector`  to determine the output data. | Required to implement at least one `transform` method |
| `void terminate(PointCollector collector) throws Exception`  | This method is called by the framework. This method will be called once after all `transform` calls have been executed. In a single UDF query, this method will and will only be called once. You need to call the data collection method provided by `collector`  to determine the output data. | Optional                                              |
| `void beforeDestroy() `                                      | This method is called by the framework after the last input data is processed, and will only be called once in the life cycle of each UDF instance. | Optional                                              |

In the life cycle of a UDTF instance, the calling sequence of each method is as follows:

1. `void validate(UDFParameterValidator validator) throws Exception`
2. `void beforeStart(UDFParameters parameters, UDTFConfigurations configurations) throws Exception`
3. `void transform(Row row, PointCollector collector) throws Exception` or `void transform(RowWindow rowWindow, PointCollector collector) throws Exception`
4. `void terminate(PointCollector collector) throws Exception`
5. `void beforeDestroy() `

Note that every time the framework executes a UDTF query, a new UDF instance will be constructed. When the query ends, the corresponding instance will be destroyed. Therefore, the internal data of the instances in different UDTF queries (even in the same SQL statement) are isolated. You can maintain some state data in the UDTF without considering the influence of concurrency and other factors.

The usage of each interface will be described in detail below.


#### void validate(UDFParameterValidator validator) throws Exception

The `validate` method is used to validate the parameters entered by the user.

In this method, you can limit the number and types of input time series, check the attributes of user input, or perform any custom verification.

Please refer to the Javadoc for the usage of `UDFParameterValidator`.


#### void beforeStart(UDFParameters parameters, UDTFConfigurations configurations) throws Exception

This method is mainly used to customize UDTF. In this method, the user can do the following things:

1. Use UDFParameters to get the time series paths and parse key-value pair attributes entered by the user.
2. Set the strategy to access the raw data and set the output data type in UDTFConfigurations.
3. Create resources, such as establishing external connections, opening files, etc.


##### UDFParameters

`UDFParameters` is used to parse UDF parameters in SQL statements (the part in parentheses after the UDF function name in SQL). The input parameters have two parts. The first part is data types of the time series that the UDF needs to process, and the second part is the key-value pair attributes for customization. Only the second part can be empty.


Example：

``` sql
SELECT UDF(s1, s2, 'key1'='iotdb', 'key2'='123.45') FROM root.sg.d;
```

Usage：

``` java
void beforeStart(UDFParameters parameters, UDTFConfigurations configurations) throws Exception {
  String stringValue = parameters.getString("key1"); // iotdb
  Float floatValue = parameters.getFloat("key2"); // 123.45
  Double doubleValue = parameters.getDouble("key3"); // null
  int intValue = parameters.getIntOrDefault("key4", 678); // 678
  // do something
  
  // configurations
  // ...
}
```


#####  UDTFConfigurations

You must use `UDTFConfigurations` to specify the strategy used by UDF to access raw data and the type of output sequence.

Usage：

``` java
void beforeStart(UDFParameters parameters, UDTFConfigurations configurations) throws Exception {
  // parameters
  // ...
  
  // configurations
  configurations
    .setAccessStrategy(new RowByRowAccessStrategy())
    .setOutputDataType(Type.INT32);
}
```

The `setAccessStrategy` method is used to set the UDF's strategy for accessing the raw data, and the `setOutputDataType` method is used to set the data type of the output sequence.


###### setAccessStrategy

Note that the raw data access strategy you set here determines which `transform` method the framework will call. Please implement the `transform` method corresponding to the raw data access strategy. Of course, you can also dynamically decide which strategy to set based on the attribute parameters parsed by `UDFParameters`. Therefore, two `transform` methods are also allowed to be implemented in one UDF.

The following are the strategies you can set:

| Interface definition              | Description                                                  | The `transform` Method to Call                               |
| :-------------------------------- | :----------------------------------------------------------- | ------------------------------------------------------------ |
| `RowByRowAccessStrategy`          | Process raw data row by row. The framework calls the `transform` method once for each row of raw data input. When UDF has only one input sequence, a row of input is one data point in the input sequence. When UDF has multiple input sequences, one row of input is a result record of the raw query (aligned by time) on these input sequences. (In a row, there may be a column with a value of `null`, but not all of them are `null`) | `void transform(Row row, PointCollector collector) throws Exception` |
| `SlidingTimeWindowAccessStrategy` | Process a batch of data in a fixed time interval each time. We call the container of a data batch a window. The framework calls the `transform` method once for each raw data input window. There may be multiple rows of data in a window, and each row is a result record of the raw query (aligned by time) on these input sequences. (In a row, there may be a column with a value of `null`, but not all of them are `null`) | `void transform(RowWindow rowWindow, PointCollector collector) throws Exception` |
| `SlidingSizeWindowAccessStrategy` | The raw data is processed batch by batch, and each batch contains a fixed number of raw data rows (except the last batch). We call the container of a data batch a window. The framework calls the `transform` method once for each raw data input window. There may be multiple rows of data in a window, and each row is a result record of the raw query (aligned by time) on these input sequences. (In a row, there may be a column with a value of `null`, but not all of them are `null`) | `void transform(RowWindow rowWindow, PointCollector collector) throws Exception` |
| `SessionTimeWindowAccessStrategy` | The raw data is processed batch by batch. We call the container of a data batch a window. The time interval between each two windows is greater than or equal to the `sessionGap` given by the user. The framework calls the `transform` method once for each raw data input window. There may be multiple rows of data in a window, and each row is a result record of the raw query (aligned by time) on these input sequences. (In a row, there may be a column with a value of `null`, but not all of them are `null`) | `void transform(RowWindow rowWindow, PointCollector collector) throws Exception` |
| `StateWindowAccessStrategy`       | The raw data is processed batch by batch. We call the container of a data batch a window. In the state window, for text type or boolean type data, each value of the point in window is equal to the value of the first point in the window, and for numerical data, the distance between each value of the point in window and the value of the first point in the window is less than the threshold `delta` given by the user. The framework calls the `transform` method once for each raw data input window. There may be multiple rows of data in a window. Currently, we only support state window for one measurement, that is, a column of data. | `void transform(RowWindow rowWindow, PointCollector collector) throws Exception` |


`RowByRowAccessStrategy`: The construction of `RowByRowAccessStrategy` does not require any parameters.

The `SlidingTimeWindowAccessStrategy` is shown schematically below.
<img style="width:100%; max-width:800px; max-height:600px; margin-left:auto; margin-right:auto; display:block;" src="https://alioss.timecho.com/docs/img/UserGuide/Process-Data/UDF-User-Defined-Function/timeWindow.png">

`SlidingTimeWindowAccessStrategy`: `SlidingTimeWindowAccessStrategy` has many constructors, you can pass 3 types of parameters to them:

- Parameter 1: The display window on the time axis
- Parameter 2: Time interval for dividing the time axis (should be positive)
- Parameter 3: Time sliding step (not required to be greater than or equal to the time interval, but must be a positive number)

The first type of parameters are optional. If the parameters are not provided, the beginning time of the display window will be set to the same as the minimum timestamp of the query result set, and the ending time of the display window will be set to the same as the maximum timestamp of the query result set.

The sliding step parameter is also optional. If the parameter is not provided, the sliding step will be set to the same as the time interval for dividing the time axis.

The relationship between the three types of parameters can be seen in the figure below. Please see the Javadoc for more details. 

<div style="text-align: center;"><img style="width:100%; max-width:800px; max-height:600px; margin-left:auto; margin-right:auto; display:block;" src="https://alioss.timecho.com/docs/img/github/99787878-47b51480-2b5b-11eb-8ed3-84088c5c30f7.png"></div>

Note that the actual time interval of some of the last time windows may be less than the specified time interval parameter. In addition, there may be cases where the number of data rows in some time windows is 0. In these cases, the framework will also call the `transform` method for the empty windows.

The `SlidingSizeWindowAccessStrategy` is shown schematically below.
<img style="width:100%; max-width:800px; max-height:600px; margin-left:auto; margin-right:auto; display:block;" src="https://alioss.timecho.com/docs/img/UserGuide/Process-Data/UDF-User-Defined-Function/countWindow.png">

`SlidingSizeWindowAccessStrategy`:  `SlidingSizeWindowAccessStrategy` has many constructors, you can pass 2 types of parameters to them:

* Parameter 1: Window size. This parameter specifies the number of data rows contained in a data processing window. Note that the number of data rows in some of the last time windows may be less than the specified number of data rows.
* Parameter 2: Sliding step. This parameter means the number of rows between the first point of the next window and the first point of the current window. (This parameter is not required to be greater than or equal to the window size, but must be a positive number)

The sliding step parameter is optional. If the parameter is not provided, the sliding step will be set to the same as the window size.

The `SessionTimeWindowAccessStrategy` is shown schematically below. **Time intervals less than or equal to the given minimum time interval `sessionGap` are assigned in one group**
<img style="width:100%; max-width:800px; max-height:600px; margin-left:auto; margin-right:auto; display:block;" src="https://alioss.timecho.com/docs/img/UserGuide/Process-Data/UDF-User-Defined-Function/sessionWindow.png">

`SessionTimeWindowAccessStrategy`: `SessionTimeWindowAccessStrategy` has many constructors, you can pass 2 types of parameters to them:

- Parameter 1: The display window on the time axis.
- Parameter 2: The minimum time interval `sessionGap` of two adjacent windows.


The `StateWindowAccessStrategy` is shown schematically below. **For numerical data, if the state difference is less than or equal to the given threshold `delta`, it will be assigned in one group. **
<img style="width:100%; max-width:800px; max-height:600px; margin-left:auto; margin-right:auto; display:block;" src="https://alioss.timecho.com/docs/img/UserGuide/Process-Data/UDF-User-Defined-Function/stateWindow.png">

`StateWindowAccessStrategy` has four constructors.

- Constructor 1: For numerical data, there are 3 parameters: the time axis can display the start and end time of the time window and the threshold `delta` for the allowable change within a single window.
- Constructor 2: For text data and boolean data, there are 3 parameters: the time axis can be provided to display the start and end time of the time window. For both data types, the data within a single window is same, and there is no need to provide an allowable change threshold.
- Constructor 3: For numerical data, there are 1 parameters: you can only provide the threshold delta that is allowed to change within a single window. The start time of the time axis display time window will be defined as the smallest timestamp in the entire query result set, and the time axis display time window end time will be defined as The largest timestamp in the entire query result set.
- Constructor 4: For text data and boolean data, you can provide no parameter. The start and end timestamps are explained in Constructor 3.

StateWindowAccessStrategy can only take one column as input for now.

Please see the Javadoc for more details. 


###### setOutputDataType

Note that the type of output sequence you set here determines the type of data that the `PointCollector` can actually receive in the `transform` method. The relationship between the output data type set in `setOutputDataType` and the actual data output type that `PointCollector` can receive is as follows:

| Output Data Type Set in `setOutputDataType` | Data Type that `PointCollector` Can Receive                  |
| :------------------------------------------ | :----------------------------------------------------------- |
| `INT32`                                     | `int`                                                        |
| `INT64`                                     | `long`                                                       |
| `FLOAT`                                     | `float`                                                      |
| `DOUBLE`                                    | `double`                                                     |
| `BOOLEAN`                                   | `boolean`                                                    |
| `TEXT`                                      | `java.lang.String` and `org.apache.iotdb.udf.api.type.Binary` |

The type of output time series of a UDTF is determined at runtime, which means that a UDTF can dynamically determine the type of output time series according to the type of input time series.
Here is a simple example:

```java
void beforeStart(UDFParameters parameters, UDTFConfigurations configurations) throws Exception {
  // do something
  // ...
  
  configurations
    .setAccessStrategy(new RowByRowAccessStrategy())
    .setOutputDataType(parameters.getDataType(0));
}
```


#### void transform(Row row, PointCollector collector) throws Exception

You need to implement this method when you specify the strategy of UDF to read the original data as `RowByRowAccessStrategy`.

This method processes the raw data one row at a time. The raw data is input from `Row` and output by `PointCollector`. You can output any number of data points in one `transform` method call. It should be noted that the type of output data points must be the same as you set in the `beforeStart` method, and the timestamps of output data points must be strictly monotonically increasing.

The following is a complete UDF example that implements the `void transform(Row row, PointCollector collector) throws Exception` method. It is an adder that receives two columns of time series as input. When two data points in a row are not `null`, this UDF will output the algebraic sum of these two data points.

``` java
import org.apache.iotdb.udf.api.UDTF;
import org.apache.iotdb.udf.api.access.Row;
import org.apache.iotdb.udf.api.collector.PointCollector;
import org.apache.iotdb.udf.api.customizer.config.UDTFConfigurations;
import org.apache.iotdb.udf.api.customizer.parameter.UDFParameters;
import org.apache.iotdb.udf.api.customizer.strategy.RowByRowAccessStrategy;
import org.apache.iotdb.udf.api.type.Type;

public class Adder implements UDTF {

  @Override
  public void beforeStart(UDFParameters parameters, UDTFConfigurations configurations) {
    configurations
        .setOutputDataType(TSDataType.INT64)
        .setAccessStrategy(new RowByRowAccessStrategy());
  }

  @Override
  public void transform(Row row, PointCollector collector) throws Exception {
    if (row.isNull(0) || row.isNull(1)) {
      return;
    }
    collector.putLong(row.getTime(), row.getLong(0) + row.getLong(1));
  }
}
```


#### void transform(RowWindow rowWindow, PointCollector collector) throws Exception

You need to implement this method when you specify the strategy of UDF to read the original data as `SlidingTimeWindowAccessStrategy` or `SlidingSizeWindowAccessStrategy`.

This method processes a batch of data in a fixed number of rows or a fixed time interval each time, and we call the container containing this batch of data a window. The raw data is input from `RowWindow` and output by `PointCollector`. `RowWindow` can help you access a batch of `Row`, it provides a set of interfaces for random access and iterative access to this batch of `Row`. You can output any number of data points in one `transform` method call. It should be noted that the type of output data points must be the same as you set in the `beforeStart` method, and the timestamps of output data points must be strictly monotonically increasing.

Below is a complete UDF example that implements the `void transform(RowWindow rowWindow, PointCollector collector) throws Exception` method. It is a counter that receives any number of time series as input, and its function is to count and output the number of data rows in each time window within a specified time range.

```java
import java.io.IOException;
import org.apache.iotdb.udf.api.UDTF;
import org.apache.iotdb.udf.api.access.Row;
import org.apache.iotdb.udf.api.access.RowWindow;
import org.apache.iotdb.udf.api.collector.PointCollector;
import org.apache.iotdb.udf.api.customizer.config.UDTFConfigurations;
import org.apache.iotdb.udf.api.customizer.parameter.UDFParameters;
import org.apache.iotdb.udf.api.customizer.strategy.SlidingTimeWindowAccessStrategy;
import org.apache.iotdb.udf.api.type.Type;

public class Counter implements UDTF {

  @Override
  public void beforeStart(UDFParameters parameters, UDTFConfigurations configurations) {
    configurations
        .setOutputDataType(TSDataType.INT32)
        .setAccessStrategy(new SlidingTimeWindowAccessStrategy(
            parameters.getLong("time_interval"),
            parameters.getLong("sliding_step"),
            parameters.getLong("display_window_begin"),
            parameters.getLong("display_window_end")));
  }

  @Override
  public void transform(RowWindow rowWindow, PointCollector collector) {
    if (rowWindow.windowSize() != 0) {
      collector.putInt(rowWindow.windowStartTime(), rowWindow.windowSize());
    }
  }
}
```


#### void terminate(PointCollector collector) throws Exception

In some scenarios, a UDF needs to traverse all the original data to calculate the final output data points. The `terminate` interface provides support for those scenarios.

This method is called after all `transform` calls are executed and before the `beforeDestory` method is executed. You can implement the `transform` method to perform pure data processing (without outputting any data points), and implement the  `terminate` method to output the processing results.

The processing results need to be output by the  `PointCollector`. You can output any number of data points in one `terminate` method call. It should be noted that the type of output data points must be the same as you set in the `beforeStart` method, and the timestamps of output data points must be strictly monotonically increasing.

Below is a complete UDF example that implements the `void terminate(PointCollector collector) throws Exception` method. It takes one time series whose data type is `INT32` as input, and outputs the maximum value point of the series.

```java
import java.io.IOException;
import org.apache.iotdb.udf.api.UDTF;
import org.apache.iotdb.udf.api.access.Row;
import org.apache.iotdb.udf.api.collector.PointCollector;
import org.apache.iotdb.udf.api.customizer.config.UDTFConfigurations;
import org.apache.iotdb.udf.api.customizer.parameter.UDFParameters;
import org.apache.iotdb.udf.api.customizer.strategy.RowByRowAccessStrategy;
import org.apache.iotdb.udf.api.type.Type;

public class Max implements UDTF {

  private Long time;
  private int value;

  @Override
  public void beforeStart(UDFParameters parameters, UDTFConfigurations configurations) {
    configurations
        .setOutputDataType(TSDataType.INT32)
        .setAccessStrategy(new RowByRowAccessStrategy());
  }

  @Override
  public void transform(Row row, PointCollector collector) {
    if (row.isNull(0)) {
      return;
    }
    int candidateValue = row.getInt(0);
    if (time == null || value < candidateValue) {
      time = row.getTime();
      value = candidateValue;
    }
  }

  @Override
  public void terminate(PointCollector collector) throws IOException {
    if (time != null) {
      collector.putInt(time, value);
    }
  }
}
```


#### void beforeDestroy() 

The method for terminating a UDF.

This method is called by the framework. For a UDF instance, `beforeDestroy` will be called after the last record is processed. In the entire life cycle of the instance, `beforeDestroy` will only be called once.


### UDAF (User Defined Aggregation Function)

A complete definition of UDAF involves two classes, `State` and `UDAF`.

#### State Class

To write your own `State`, you need to implement the `org.apache.iotdb.udf.api.State` interface.

The following table shows all the interfaces available for user implementation.

| Interface Definition             | Description                                                  | Required to Implement |
| -------------------------------- | ------------------------------------------------------------ | --------------------- |
| `void reset()`                   | To reset the `State` object to its initial state, you need to fill in the initial values of the fields in the `State` class within this method as if you were writing a constructor. | Required              |
| `byte[] serialize()`             | Serializes `State` to binary data. This method is used for IoTDB internal `State` passing. Note that the order of serialization must be consistent with the following deserialization methods. | Required              |
| `void deserialize(byte[] bytes)` | Deserializes binary data to `State`. This method is used for IoTDB internal  `State` passing. Note that the order of deserialization must be consistent with the serialization method above. | Required              |

The following section describes the usage of each interface in detail.


##### void reset()

This method resets the `State` to its initial state, you need to fill in the initial values of the fields in the `State` object in this method. For optimization reasons, IoTDB reuses `State` as much as possible internally, rather than creating a new `State` for each group, which would introduce unnecessary overhead. When `State` has finished updating the data in a group, this method is called to reset to the initial state as a way to process the next group.

In the case of `State` for averaging (aka `avg`), for example, you would need the sum of the data, `sum`, and the number of entries in the data, `count`, and initialize both to 0 in the `reset()` method.

```java
class AvgState implements State {
  double sum;

  long count;

  @Override
  public void reset() {
    sum = 0;
    count = 0;
  }
  
  // other methods
}
```


##### byte[] serialize()/void deserialize(byte[] bytes)

These methods serialize the `State` into binary data, and deserialize the `State` from the binary data. IoTDB, as a distributed database, involves passing data among different nodes, so you need to write these two methods to enable the passing of the State among different nodes. Note that the order of serialization and deserialization must be the consistent.

In the case of `State` for averaging (aka `avg`), for example, you can convert the content of State to `byte[]` array and read out the content of State from `byte[]` array in any way you want, the following shows the code for serialization/deserialization using `ByteBuffer` introduced by Java8:

```java
@Override
public byte[] serialize() {
  ByteBuffer buffer = ByteBuffer.allocate(Double.BYTES + Long.BYTES);
  buffer.putDouble(sum);
  buffer.putLong(count);

  return buffer.array();
}

@Override
public void deserialize(byte[] bytes) {
  ByteBuffer buffer = ByteBuffer.wrap(bytes);
  sum = buffer.getDouble();
  count = buffer.getLong();
}
```


#### UDAF Classes

To write a UDAF, you need to implement the `org.apache.iotdb.udf.api.UDAF` interface.

The following table shows all the interfaces available for user implementation.

| Interface definition                                         | Description                                                  | Required to Implement |
| ------------------------------------------------------------ | ------------------------------------------------------------ | --------------------- |
| `void validate(UDFParameterValidator validator) throws Exception` | This method is mainly used to validate `UDFParameters` and it is executed before `beforeStart(UDFParameters, UDTFConfigurations)` is called. | Optional              |
| `void beforeStart(UDFParameters parameters, UDAFConfigurations configurations) throws Exception` | Initialization method that invokes user-defined initialization behavior before UDAF processes the input data. Unlike UDTF, configuration is of type `UDAFConfiguration`. | Required              |
| `State createState()`                                        | To create a `State` object, usually just call the default constructor and modify the default initial value as needed. | Required              |
| `void addInput(State state, Column[] columns, BitMap bitMap)` | Update `State` object according to the incoming data `Column[]` in batch, note that `column[0]` always represents the time column. In addition, `BitMap` represents the data that has been filtered out before, you need to manually determine whether the corresponding data has been filtered out when writing this method. | Required              |
| `void combineState(State state, State rhs)`                  | Merge `rhs` state into `state` state. In a distributed scenario, the same set of data may be distributed on different nodes, IoTDB generates a `State` object for the partial data on each node, and then calls this method to merge it into the complete `State`. | Required              |
| `void outputFinal(State state, ResultValue resultValue)`     | Computes the final aggregated result based on the data in `State`. Note that according to the semantics of the aggregation, only one value can be output per group. | Required              |
| `void beforeDestroy() `                                      | This method is called by the framework after the last input data is processed, and will only be called once in the life cycle of each UDF instance. | Optional              |

In the life cycle of a UDAF instance, the calling sequence of each method is as follows:

1. `State createState()`
2. `void validate(UDFParameterValidator validator) throws Exception`
3. `void beforeStart(UDFParameters parameters, UDAFConfigurations configurations) throws Exception`
4. `void addInput(State state, Column[] columns, BitMap bitMap)` 
5. `void combineState(State state, State rhs)` 
6. `void outputFinal(State state, ResultValue resultValue)` 
7. `void beforeDestroy()`

Similar to UDTF, every time the framework executes a UDAF query, a new UDF instance will be constructed. When the query ends, the corresponding instance will be destroyed. Therefore, the internal data of the instances in different UDAF queries (even in the same SQL statement) are isolated. You can maintain some state data in the UDAF without considering the influence of concurrency and other factors.

The usage of each interface will be described in detail below.


##### void validate(UDFParameterValidator validator) throws Exception

Same as UDTF, the `validate` method is used to validate the parameters entered by the user.

In this method, you can limit the number and types of input time series, check the attributes of user input, or perform any custom verification.


##### void beforeStart(UDFParameters parameters, UDAFConfigurations configurations) throws Exception

 The `beforeStart` method does the same thing as the UDAF:

1. Use UDFParameters to get the time series paths and parse key-value pair attributes entered by the user.
2. Set the strategy to access the raw data and set the output data type in UDAFConfigurations.
3. Create resources, such as establishing external connections, opening files, etc.

The role of the `UDFParameters` type can be seen above.

###### UDAFConfigurations

The difference from UDTF is that UDAF uses `UDAFConfigurations` as the type of `configuration` object.

Currently, this class only supports setting the type of output data.

```java
void beforeStart(UDFParameters parameters, UDAFConfigurations configurations) throws Exception {
  // parameters
  // ...

  // configurations
  configurations
    .setOutputDataType(Type.INT32); }
}
```

The relationship between the output type set in `setOutputDataType` and the type of data output that `ResultValue` can actually receive is as follows:

| The output type set in `setOutputDataType` | The output type that `ResultValue` can actually receive |
| ------------------------------------------ | ------------------------------------------------------- |
| `INT32`                                    | `int`                                                   |
| `INT64`                                    | `long`                                                  |
| `FLOAT`                                    | `float`                                                 |
| `DOUBLE`                                   | `double`                                                |
| `BOOLEAN`                                  | `boolean`                                               |
| `TEXT`                                     | `org.apache.iotdb.udf.api.type.Binary`                  |

The output type of the UDAF is determined at runtime. You can dynamically determine the output sequence type based on the input type.

Here is a simple example:

```java
void beforeStart(UDFParameters parameters, UDAFConfigurations configurations) throws Exception {
  // do something
  // ...
  
  configurations
    .setOutputDataType(parameters.getDataType(0));
}
```


##### State createState()

This method creates and initializes a `State` object for UDAF. Due to the limitations of the Java language, you can only call the default constructor for the `State` class. The default constructor assigns a default initial value to all the fields in the class, and if that initial value does not meet your requirements, you need to initialize them manually within this method.

The following is an example that includes manual initialization. Suppose you want to implement an aggregate function that multiply all numbers in the group, then your initial `State` value should be set to 1, but the default constructor initializes it to 0, so you need to initialize `State` manually after calling the default constructor:

```java
public State createState() {
  MultiplyState state = new MultiplyState();
  state.result = 1;
  return state;
}
```


##### void addInput(State state, Column[] columns, BitMap bitMap)

This method updates the `State` object with the raw input data. For performance reasons, also to align with the IoTDB vectorized query engine, the raw input data is no longer a data point, but an array of columns ``Column[]``. Note that the first column (i.e. `column[0]`) is always the time column, so you can also do different operations in UDAF depending on the time.

Since the input parameter is not of a single data point type, but of multiple columns, you need to manually filter some of the data in the columns, which is why the third parameter, `BitMap`, exists. It identifies which of these columns have been filtered out, so you don't have to think about the filtered data in any case.

Here's an example of `addInput()` that counts the number of items (aka count). It shows how you can use `BitMap` to ignore data that has been filtered out. Note that due to the limitations of the Java language, you need to do the explicit cast the `State` object  from type defined in the interface to a custom `State` type at the beginning of the method, otherwise you won't be able to use the `State` object.

```java
public void addInput(State state, Column[] column, BitMap bitMap) {
  CountState countState = (CountState) state;

  int count = column[0].getPositionCount();
  for (int i = 0; i < count; i++) {
    if (bitMap != null && !bitMap.isMarked(i)) {
      continue;
    }
    if (!column[1].isNull(i)) {
      countState.count++;
    }
  }
}
```


##### void combineState(State state, State rhs)

This method combines two `State`s, or more precisely, updates the first `State` object with the second `State` object. IoTDB is a distributed database, and the data of the same group may be distributed on different nodes. For performance reasons, IoTDB will first aggregate some of the data on each node into `State`, and then merge the `State`s on different nodes that belong to the same group, which is what `combineState` does.

Here's an example of `combineState()` for averaging (aka avg). Similar to `addInput`, you need to do an explicit type conversion for the two `State`s at the beginning. Also note that you are updating the value of the first `State` with the contents of the second `State`.

```java
public void combineState(State state, State rhs) {
  AvgState avgState = (AvgState) state;
  AvgState avgRhs = (AvgState) rhs;

  avgState.count += avgRhs.count;
  avgState.sum += avgRhs.sum;
}
```


##### void outputFinal(State state, ResultValue resultValue)

This method works by calculating the final result from `State`. You need to access the various fields in `State`, derive the final result, and set the final result into the `ResultValue` object.IoTDB internally calls this method once at the end for each group. Note that according to the semantics of aggregation, the final result can only be one value.

Here is another `outputFinal` example for averaging (aka avg). In addition to the forced type conversion at the beginning, you will also see a specific use of the `ResultValue` object, where the final result is set by `setXXX` (where `XXX` is the type name).

```java
public void outputFinal(State state, ResultValue resultValue) {
  AvgState avgState = (AvgState) state;

  if (avgState.count != 0) {
    resultValue.setDouble(avgState.sum / avgState.count);
  } else {
    resultValue.setNull();
  }
}
```


##### void beforeDestroy()

The method for terminating a UDF.

This method is called by the framework. For a UDF instance, `beforeDestroy` will be called after the last record is processed. In the entire life cycle of the instance, `beforeDestroy` will only be called once.


### Maven Project Example

If you use Maven, you can build your own UDF project referring to our **udf-example** module. You can find the project [here](https://github.com/apache/iotdb/tree/master/example/udf).


### UDF Registration

The process of registering a UDF in IoTDB is as follows:

1. Implement a complete UDF class, assuming the full class name of this class is `org.apache.iotdb.udf.ExampleUDTF`.
2. Package your project into a JAR. If you use Maven to manage your project, you can refer to the Maven project example above.
3. Make preparations for registration according to the registration mode. For details, see the following example.
4. You can use following SQL to register UDF.

```sql
CREATE FUNCTION <UDF-NAME> AS <UDF-CLASS-FULL-PATHNAME> (USING URI URI-STRING)?
```

#### Example: register UDF named `example`, you can choose either of the following two registration methods

##### No URI

Prepare:  
When use this method to register，you should put JAR to directory `iotdb-server-1.0.0-all-bin/ext/udf`（directory can config）.  
**Note，you should put JAR to this directory of all DataNodes if using Cluster**

SQL:  

```sql
CREATE FUNCTION example AS 'org.apache.iotdb.udf.UDTFExample'
```

##### Using URI

Prepare:  
When use this method to register，you need to upload the JAR to URI server and ensure the IoTDB instance executing this registration statement has access to the URI server.  
**Note，you needn't place JAR manually，IoTDB will download the JAR and sync it.**

SQL:

```sql
CREATE FUNCTION example AS 'org.apache.iotdb.udf.UDTFExample' USING URI 'http://jar/example.jar'
```

#### Note

Since UDF instances are dynamically loaded through reflection technology, you do not need to restart the server during the UDF registration process.

UDF function names are not case-sensitive.

Please ensure that the function name given to the UDF is different from all built-in function names. A UDF with the same name as a built-in function cannot be registered.

We recommend that you do not use classes that have the same class name but different function logic in different JAR packages. For example, in `UDF(UDAF/UDTF): udf1, udf2`, the JAR package of udf1 is `udf1.jar` and the JAR package of udf2 is `udf2.jar`. Assume that both JAR packages contain the `org.apache.iotdb.udf.ExampleUDTF` class. If you use two UDFs in the same SQL statement at the same time, the system will randomly load either of them and may cause inconsistency in UDF execution behavior.

### UDF Deregistration

The following shows the SQL syntax of how to deregister a UDF.

```sql
DROP FUNCTION <UDF-NAME>
```

Here is an example:

```sql
DROP FUNCTION example
```


### UDF Queries

The usage of UDF is similar to that of built-in aggregation functions.


#### Basic SQL syntax support

* Support `SLIMIT` / `SOFFSET`
* Support `LIMIT` / `OFFSET`
* Support queries with time filters
* Support queries with value filters


#### Queries with * in SELECT Clauses

Assume that there are 2 time series (`root.sg.d1.s1` and `root.sg.d1.s2`)  in the system.

* **`SELECT example(*) from root.sg.d1`**

Then the result set will include the results of `example (root.sg.d1.s1)` and `example (root.sg.d1.s2)`.

* **`SELECT example(s1, *) from root.sg.d1`**

Then the result set will include the results of `example(root.sg.d1.s1, root.sg.d1.s1)` and `example(root.sg.d1.s1, root.sg.d1.s2)`.

* **`SELECT example(*, *) from root.sg.d1`**

Then the result set will include the results of  `example(root.sg.d1.s1, root.sg.d1.s1)`, `example(root.sg.d1.s2, root.sg.d1.s1)`, `example(root.sg.d1.s1, root.sg.d1.s2)` and `example(root.sg.d1.s2, root.sg.d1.s2)`.


#### Queries with Key-value Attributes in UDF Parameters

You can pass any number of key-value pair parameters to the UDF when constructing a UDF query. The key and value in the key-value pair need to be enclosed in single or double quotes. Note that key-value pair parameters can only be passed in after all time series have been passed in. Here is a set of examples:

``` sql
SELECT example(s1, 'key1'='value1', 'key2'='value2'), example(*, 'key3'='value3') FROM root.sg.d1;
SELECT example(s1, s2, 'key1'='value1', 'key2'='value2') FROM root.sg.d1;
```


#### Nested Queries

``` sql
SELECT s1, s2, example(s1, s2) FROM root.sg.d1;
SELECT *, example(*) FROM root.sg.d1 DISABLE ALIGN;
SELECT s1 * example(* / s1 + s2) FROM root.sg.d1;
SELECT s1, s2, s1 + example(s1, s2), s1 - example(s1 + example(s1, s2) / s2) FROM root.sg.d1;
```


### Show All Registered UDFs

``` sql
SHOW FUNCTIONS
```


### User Permission Management

There are 1 types of user permissions related to UDF: `USE_UDF`

* Only users with this permission are allowed to register UDFs
* Only users with this permission are allowed to deregister UDFs
* Only users with this permission are allowed to use UDFs for queries

For more user permissions related content, please refer to [Account Management Statements](./Authority-Management.md).


### Configurable Properties

You can use `udf_lib_dir` to config udf lib directory.  
When querying by a UDF, IoTDB may prompt that there is insufficient memory. You can resolve the issue by configuring `udf_initial_byte_array_length_for_memory_control`, `udf_memory_budget_in_mb` and `udf_reader_transformer_collector_memory_proportion` in `iotdb-datanode.properties` and restarting the server.


### Contribute UDF

<!-- The template is copied and modified from the Apache Doris community-->

This part mainly introduces how external users can contribute their own UDFs to the IoTDB community.


#### Prerequisites

1. UDFs must be universal.

    The "universal" mentioned here refers to: UDFs can be widely used in some scenarios. In other words, the UDF function must have reuse value and may be directly used by other users in the community.

    If you are not sure whether the UDF you want to contribute is universal, you can send an email to `dev@iotdb.apache.org` or create an issue to initiate a discussion.

2. The UDF you are going to contribute has been well tested and can run normally in the production environment.


#### What you need to prepare

1. UDF source code
2. Test cases
3. Instructions


##### UDF Source Code

1. Create the UDF main class and related classes in `iotdb-core/node-commons/src/main/java/org/apache/iotdb/commons/udf/builtin` or in its subfolders.
2. Register your UDF in `iotdb-core/node-commons/src/main/java/org/apache/iotdb/commons/udf/builtin/BuiltinTimeSeriesGeneratingFunction.java`.


##### Test Cases

At a minimum, you need to write integration tests for the UDF.

You can add a test class in `integration-test/src/test/java/org/apache/iotdb/db/it/udf`. 


##### Instructions

The instructions need to include: the name and the function of the UDF, the attribute parameters that must be provided when the UDF is executed, the applicable scenarios, and the usage examples, etc.

The instructions should be added in `docs/UserGuide/Operation Manual/DML Data Manipulation Language.md`.


#### Submit a PR

When you have prepared the UDF source code, test cases, and instructions, you are ready to submit a Pull Request (PR) on [Github](https://github.com/apache/iotdb). You can refer to our code contribution guide to submit a PR: [Pull Request Guide](https://iotdb.apache.org/Development/HowToCommit.html).

### Known Implementations

#### Built-in UDF

1.   Aggregate Functions, such as `SUM`. For details and examples, see the document [Aggregate Functions](../Reference/Function-and-Expression.md#aggregate-functions).
2.   Arithmetic Functions, such as `SIN`. For details and examples, see the document [Arithmetic Operators and Functions](../Reference/Function-and-Expression.md#arithmetic-operators-and-functions).
3.   Comparison Functions, such as `ON_OFF`. For details and examples, see the document [Comparison Operators and Functions](../Reference/Function-and-Expression.md#comparison-operators-and-functions).
4.   String Processing Functions, such as `STRING_CONTAINS`. For details and examples, see the document [String Processing](../Reference/Function-and-Expression.md#string-processing).
5.   Data Type Conversion Function, such as `CAST`. For details and examples, see the document [Data Type Conversion Function](../Reference/Function-and-Expression.md#data-type-conversion-function).
6.   Constant Timeseries Generating Functions, such as `CONST`. For details and examples, see the document [Constant Timeseries Generating Functions](../Reference/Function-and-Expression.md#constant-timeseries-generating-functions).
7.   Selector Functions, such as `TOP_K`. For details and examples, see the document [Selector Functions](../Reference/Function-and-Expression.md#selector-functions).
8.   Continuous Interval Functions, such as `ZERO_DURATION`. For details and examples, see the document [Continuous Interval Functions](../Reference/Function-and-Expression.md#continuous-interval-functions).
9.   Variation Trend Calculation Functions, such as `TIME_DIFFERENCE`. For details and examples, see the document [Variation Trend Calculation Functions](../Reference/Function-and-Expression.md#variation-trend-calculation-functions).
10.   Sample Functions, such as `M4`. For details and examples, see the document [Sample Functions](../Reference/Function-and-Expression.md#sample-functions).
11.   Change Points Function, such as `CHANGE_POINTS`. For details and examples, see the document [Time-Series](../Reference/Function-and-Expression.md#time-series-processing).

#### Data Quality Function Library

##### About

For applications based on time series data, data quality is vital. **UDF Library** is IoTDB User Defined Functions (UDF) about data quality, including data profiling, data quality evalution and data repairing. It effectively meets the demand for data quality in the industrial field.

##### Quick Start

The functions in this function library are not built-in functions, and must be loaded into the system before use.

1. [Download](https://archive.apache.org/dist/iotdb/1.0.1/apache-iotdb-1.0.1-library-udf-bin.zip) the JAR with all dependencies and the script of registering UDF.
2. Copy the JAR package to `ext\udf` under the directory of IoTDB system (Please put JAR to this directory of all DataNodes if you use Cluster).
3. Run `sbin\start-server.bat` (for Windows) or `sbin\start-server.sh` (for Linux or MacOS) to start IoTDB server.
4. Copy the script to the directory of IoTDB system (under the root directory, at the same level as `sbin`), modify the parameters in the script if needed and run it to register UDF.

##### Implemented Functions

1.   Data Quality related functions, such as `Completeness`. For details and examples, see the document [Data-Quality](../Reference/UDF-Libraries.md#data-quality).
2.   Data Profiling related functions, such as `ACF`. For details and examples, see the document [Data-Profiling](../Reference/UDF-Libraries.md#data-profiling).
3.   Anomaly Detection related functions, such as `IQR`. For details and examples, see the document [Anomaly-Detection](../Reference/UDF-Libraries.md#anomaly-detection).
4.   Frequency Domain Analysis related functions, such as `Conv`. For details and examples, see the document [Frequency-Domain](../Reference/UDF-Libraries.md#frequency-domain-analysis).
5.   Data Matching related functions, such as `DTW`. For details and examples, see the document [Data-Matching](../Reference/UDF-Libraries.md#data-matching).
6.   Data Repairing related functions, such as `TimestampRepair`. For details and examples, see the document [Data-Repairing](../Reference/UDF-Libraries.md#data-repairing).
7.   Series Discovery related functions, such as `ConsecutiveSequences`. For details and examples, see the document [Series-Discovery](../Reference/UDF-Libraries.md#series-discovery).
8.   Machine Learning related functions, such as `AR`. For details and examples, see the document [Machine-Learning](../Reference/UDF-Libraries.md#machine-learning).


### Q&A

Q1: How to modify the registered UDF? 

A1: Assume that the name of the UDF is `example` and the full class name is `org.apache.iotdb.udf.ExampleUDTF`, which is introduced by `example.jar`.

1. Unload the registered function by executing `DROP FUNCTION example`.
2. Delete `example.jar` under `iotdb-server-1.0.0-all-bin/ext/udf`.
3. Modify the logic in `org.apache.iotdb.udf.ExampleUDTF` and repackage it. The name of the JAR package can still be `example.jar`.
4. Upload the new JAR package to `iotdb-server-1.0.0-all-bin/ext/udf`.
5. Load the new UDF by executing `CREATE FUNCTION example AS "org.apache.iotdb.udf.ExampleUDTF"`.