.. _auto_dataset_compute_prediction_accuracy:

*DATASET COMPUTE PREDICTION ACCURACY*
=====================================

Computes prediction accuracy for a set of new predictions by comparing them against training data using either a ratio or a normalized maximum error metric. Use this worker after model inference to quantify how well new predictions match the original target values.

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

Classification: **process**.

Tagged: ``error_estimation``, ``model_evaluation``, ``normalized_error``, ``prediction_accuracy``, ``ratio_error``, ``regression_metrics``.

Inputs
------

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

   * - Label
     - ID
     - Type
     - Default
     - Required
     - Description
   * - Dataset
     - dataset_1
     - dataset
     - —
     - 
     - Training dataset containing the original input and target columns used to build the model; serves as the reference ground-truth for accuracy comparison.
   * - Input cols
     - input_cols
     - scalar
     - —
     - ✓
     - Comma-separated list of column names from dataset_1 that were used as model input/feature columns during training; must match the column names exactly.
   * - Target cols
     - target_cols
     - scalar
     - —
     - ✓
     - Comma-separated list of column names from dataset_1 that were used as model target/output columns during training; must match the column names exactly.
   * - New Predictions
     - new_predictions
     - dataset
     - —
     - ✓
     - Dataset of new model predictions whose accuracy is to be estimated; must contain the same input and target column structure as dataset_1.
   * - Error Calculation Type
     - calc_type
     - string
     - 1
     - 
     - Error calculation method: '1' computes the ratio of predicted to raw (actual) value; '2' computes the normalized maximum error between predicted and raw values; defaults to '1' (ratio).

Outputs
-------

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

   * - Label
     - ID
     - Type
     - Description
   * - New Predictions with Accuracy
     - dataset_prediction_accuracy
     - dataset
     - Augmented version of the new predictions dataset with appended accuracy/error columns for each target, reflecting the chosen error metric (ratio or normalized max error) per prediction row.

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

- ai_ml.prognosis
- ai_ml.supervised.regression
- 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_compute_prediction_accuracy <https://www.d3view.com/api/workflow/example?id=dataset_compute_prediction_accuracy>`_

.. raw:: html

   <hr style="margin-top:2em">
   <p style="font-size:11px;color:#888">
   Auto-generated from <code>transformation</code> schema. Worker id: <code>dataset_compute_prediction_accuracy</code>. Schema hash: <code>6b8035375882</code>. Hand-curated docs in <code>workerexamples/</code> override this page when present.
   </p>