Enhance Documentation for Dataset Previews

docs/source/preview_custom_datasets.md

@@ -11,13 +11,20 @@
 JSONPreview = NewType("JSONPreview", dict)
 ```
 
-Arbitrary arguments can be included in the `preview()` function, which can be later specified in the `catalog.yml` file.
+## TablePreview
+For `TablePreview`, the returned dictionary must contain the following keys:
+
+`index`: A list of row indices.
+`columns`: A list of column names.
+`data`: A list of rows, where each row is itself a list of values.
+
+Arbitrary arguments can be included in the `preview()` function, which can be later specified in the `catalog.yml` file. Ensure that these arguments (like `nrows`, `ncolumns`, and `filters`) match the structure of your dataset.
 
 Below is an example demonstrating how to implement the `preview()` function with user-specified arguments for a `CustomDataset` class that utilizes `TablePreview` to enable previewing tabular data on Kedro-Viz:
 
 ```yaml 
 companies:
-  type: CustomDataset
+  type: CustomTableDataset
   filepath: ${_base_location}/01_raw/companies.csv
   metadata:
     kedro-viz:
@@ -34,36 +41,73 @@
 
 from kedro_datasets._typing import TablePreview
 
-class CustomDataset:
+class CustomTableDataset:
   def preview(self, nrows, ncolumns, filters) -> TablePreview:
-    filtered_data = self.data
+    data = self.load()
     for column, value in filters.items():
-        filtered_data = filtered_data[filtered_data[column] == value]
-    subset = filtered_data.iloc[:nrows, :ncolumns]
-    df_dict = {}
-    for column in subset.columns:
-        df_dict[column] = subset[column]
-    return df_dict
-
+        data = data[data[column] == value]
+    subset = data.iloc[:nrows, :ncolumns]
+    preview_data = {
+        'index': list(subset.index),  # List of row indices
+        'columns': list(subset.columns),  # List of column names
+        'data': subset.values.tolist()  # List of rows, where each row is a list of values
+    }
+    return preview_data
 ```
 
+![](./images/preview_datasets_expanded.png)
 
-## Examples of Previews
+## ImagePreview
+For `ImagePreview`, the function should return a base64-encoded string representing the image. This is typically used for datasets that output visual data such as plots or images.
 
-1. TablePreview 
+Below is an example implementation:
 
-![](./images/preview_datasets_expanded.png)
+```python
 
+from kedro_datasets._typing import ImagePreview
 
-2. ImagePreview
+class CustomImageDataset:
+  def preview(self) -> ImagePreview:
+    image_path = self._get_image_path()
+    with open(image_path, "rb") as image_file:
+        encoded_string = base64.b64encode(image_file.read()).decode('utf-8')
+    return ImagePreview(encoded_string)
+```
 
 ![](./images/pipeline_visualisation_matplotlib_expand.png)
 
+## PlotlyPreview
+For `PlotlyPreview`, the function should return a dictionary containing Plotly figure data. This includes the figure's `data` and `layout` keys.
+
+Below is an example implementation:
 
-3. PlotlyPreview
+```python
+
+from kedro_datasets._typing import PlotlyPreview
+
+class CustomPlotlyDataset:
+  def preview(self) -> PlotlyPreview:
+    figure = self._load_plotly_figure()
+    return PlotlyPreview({
+        "data": figure["data"],
+        "layout": figure["layout"]
+    })
+```
 
 ![](./images/pipeline_visualisation_plotly_expand_1.png)
 
+## JSONPreview
+For `JSONPreview`, the function should return a dictionary representing the `JSON` data. This is useful for previewing complex nested data structures.
 
+Below is an example implementation:
 
+```python
+
+from kedro_datasets._typing import JSONPreview
 
+class CustomJSONDataset:
+  def preview(self) -> JSONPreview:
+    json_data = self._load_json_data()
+    return JSONPreview(json.dumps(json_data))
+```
+![](./images/preview_datasets_json.png)

docs/source/preview_datasets.md

@@ -57,6 +57,8 @@ companies:
       preview: false
 ```
 
+You can also disable previews globally through the settings menu on Kedro-Viz.
+
 ```{note}
 Starting from Kedro-Viz 9.2.0, previews are disabled by default for the CLI commands `kedro viz deploy` and `kedro viz build`. You can control this behavior using the `--include-previews` flag with these commands. For `kedro viz run`, previews are enabled by default and can be controlled from the publish modal dialog, refer to the [Publish and share](./share_kedro_viz) for more instructions.
 ```