blob: 07d83181d77e15f2d359893dc140b2e5397e3233 [file]
Density Sketch
--------------
.. currentmodule:: datasketches
Builds a coreset from the given set of input points.
Provides density estimate at a given point.
Based on the following paper: Zohar Karnin, Edo Liberty
"Discrepancy, Coresets, and Sketches in Machine Learning"
https://proceedings.mlr.press/v99/karnin19a/karnin19a.pdf
Inspired by the following implementation: https://github.com/edoliberty/streaming-quantiles/blob/f688c8161a25582457b0a09deb4630a81406293b/gde.py
Requires the use of a :class:`KernelFunction` to compute the distance between two vectors.
.. autoclass:: density_sketch
:members:
:undoc-members:
:exclude-members: deserialize
.. rubric:: Static Methods:
.. automethod:: deserialize
.. rubric:: Non-static Methods:
.. automethod:: __init__