Add document descriptions for Python APIs (#761)

* add document descriptions for python api

* modify CHANGELOG

* remark-lint line length 100

* update CHANGELOG

* correct whitespace

CHANGELOG.md

@@ -25,8 +25,10 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
     (<https://github.com/openvinotoolkit/datumaro/pull/746>)
   - Add Mask, SuperResolution, Depth visualization features
     (<https://github.com/openvinotoolkit/datumaro/pull/747>)
-- Add a documentation tab menu for Python API
+- Add a documentation for Python API
   (<https://github.com/openvinotoolkit/datumaro/pull/753>)
+  - Add dataset handler, visualizer, filter descriptions
+    (<https://github.com/openvinotoolkit/datumaro/pull/761>)
 - Add `__repr__` for Dataset
   (<https://github.com/openvinotoolkit/datumaro/pull/750>)
 - Support for exporting as CVAT video format

site/content/en/docs/python-api/_index.md

@@ -10,6 +10,7 @@ Need to update the description.
 
 ## Contents
 - [Python API examples](./python-api-examples)
+  - [Dataset Handler](./python-api-examples/dataset-handler)
   - [Visualizer](./python-api-examples/visualizer)
   - [Filter](./python-api-examples/filter)
   - [Transform](./python-api-examples/transform)

site/content/en/docs/python-api/python-api-examples/dataset_handler.md

@@ -0,0 +1,16 @@
+---
+title: 'Dataset handler'
+linkTitle: 'dataset-handler'
+description: ''
+---
+
+Datumaro provides the dataset import and export functionalities.
+
+When importing multiple datasets, Datumaro helps to manipulate and merge them into a single
+dataset. Since the manipulations such as reidentification, label redefinition, or filtration are
+mostly the topic of transformation, we here describe how to merge two heterogeneous datasets
+through `IntersectMerge`.
+
+Jupyter Notebook Examples:
+{{< blocks/notebook 01_merge_multiple_datasets_for_classification >}}
+{{< blocks/notebook 02_merge_heterogeneous_datasets_for_detection >}}

site/content/en/docs/python-api/python-api-examples/filter.md

@@ -4,7 +4,50 @@ linkTitle: 'filter'
 description: ''
 ---
 
-Need to update the description.
+This API allows you to filter a dataset to satisfy some conditions.
+Here, XML [XPath](https://devhints.io/xpath) is used as a query format.
 
-Jupyter Notebook Examples:
-Need to update the description.
+For instance, with a given XML file below, we can filter a dataset by the subset name through
+`/item[subset="minival2014"]`, by the media id through `/item[id="290768"]`, by the image sizes
+through `/item[image/width=image/height]`, and annotation information such as id (`id`), type
+(`type`), label (`label_id`), bounding box (`x, y, w, h`), etc.
+
+``` xml
+<item>
+  <id>290768</id>
+  <subset>minival2014</subset>
+  <image>
+    <width>612</width>
+    <height>612</height>
+    <depth>3</depth>
+  </image>
+  <annotation>
+    <id>80154</id>
+    <type>bbox</type>
+    <label_id>39</label_id>
+    <x>264.59</x>
+    <y>150.25</y>
+    <w>11.19</w>
+    <h>42.31</h>
+    <area>473.87</area>
+  </annotation>
+  <annotation>
+    <id>669839</id>
+    <type>bbox</type>
+    <label_id>41</label_id>
+    <x>163.58</x>
+    <y>191.75</y>
+    <w>76.98</w>
+    <h>73.63</h>
+    <area>5668.77</area>
+  </annotation>
+  ...
+</item>
+```
+
+For the annotation-based filtration, we need to set the argument `filter_annotations` to `True`.
+We provide the argument `remove_empty` to remove all media with an empty annotation. We note that
+datasets are updated in-place by default.
+
+Jupyter Notebook Example:
+{{< blocks/notebook 04_filter >}}

site/content/en/docs/python-api/python-api-examples/visualizer.md

@@ -4,7 +4,13 @@ linkTitle: 'visualizer'
 description: ''
 ---
 
-Need to update the description.
+This API allows you to visualize a dataset with media ids.
 
-Jupyter Notebook Examples:
+Although Datumaro supports various kinds of vision tasks, e.g., classification, object detection,
+semantic segmentation, key point estimation, visual captioning, etc., we provide a task-agnostic
+visualization tool. That is, regardless of annotation types, `vis_gallery` describes the multiple
+annotation-overlapped images from a list of multiple media ids. We can control the transparency of
+annotations over images by adjusting `alpha`.
+
+Jupyter Notebook Example:
 {{< blocks/notebook 03_visualize >}}