.. _auto_dataset_analytics:

*DATASET ANALYTICS*
===================

Analyzes a tabular dataset using one of six configurable methods (summary statistics, value counts, group-by aggregate, correlation, trend, or outlier detection). Analysis parameters can be supplied as a raw JSON config or generated automatically via a natural-language prompt. Use this worker to perform exploratory data analysis or derive KPIs from any workflow dataset.

When to use
-----------

Tagged: ``ai_configuration``, ``analytics``, ``correlation``, ``dataset``, ``eda``, ``group_by``, ``llm_config``, ``outlier_detection``.

Inputs
------

.. list-table::
   :header-rows: 1
   :widths: 20 20 20 20 20 20

   * - Label
     - ID
     - Type
     - Default
     - Required
     - Description
   * - Dataset
     - dataset
     - dataset
     - —
     - ✓
     - Tabular dataset to analyze, supplied as an array of records (rows); must be non-empty and contain at least one valid row after cleaning.
   * - Method
     - method
     - select
     - —
     - ✓
     - Analysis method to apply; choose one of: summary_statistics, value_counts, group_by_aggregate, correlation_analysis, trend_analysis, or outlier_detection.
   * - Custom Configs
     - custom_configs
     - textarea
     - —
     - 
     - Optional JSON string containing method-specific configuration parameters (e.g., column names, aggregation functions, thresholds); required when no prompt is provided; leave blank to use AI-generated config instead.
   * - Prompt
     - prompt
     - textarea
     - —
     - 
     - Optional natural-language prompt describing the desired analysis; used to auto-generate method configs via an AI configuration worker when custom_configs is not supplied.

Outputs
-------

.. list-table::
   :header-rows: 1
   :widths: 20 20 20 20

   * - Label
     - ID
     - Type
     - Description
   * - Results
     - results
     - string
     - String-serialised analysis output whose structure depends on the chosen method (e.g., descriptive stats table, correlation matrix, outlier row list); contains an error message string on failure.
   * - Status
     - status
     - text
     - JSON-encoded status object with keys success (bool), message (string), original_rows, excluded_rows, and final_rows; indicates whether the analysis completed successfully and summarises dataset cleaning counts.

Disciplines
-----------

- ai_ml.llm
- ai_ml.unsupervised.anomaly_detection
- data.dataset.transform
- data.statistics

Runnable example
----------------

A runnable example is registered for this worker. Open the example workflow on the d3VIEW canvas: `/api/workflow/example?id=dataset_analytics <https://www.d3view.com/api/workflow/example?id=dataset_analytics>`_

.. raw:: html

   <hr style="margin-top:2em">
   <p style="font-size:11px;color:#888">
   Auto-generated from <code>platform</code> schema. Worker id: <code>dataset_analytics</code>. Schema hash: <code>da8eb166e936</code>. Hand-curated docs in <code>workerexamples/</code> override this page when present.
   </p>