.. _auto_dataset_compute_derivatives:

*DATASET COMPUTE DERIVATIVES*
=============================

Computes numerical derivatives (dy/dx) of one or more target columns with respect to one or more input columns in a dataset. Optionally builds a Full Factorial Design (FFD) dataset from the source data before differentiation, and supports multiple neighbor-based derivative estimation strategies. Use this worker to enrich a design-exploration dataset with gradient information for sensitivity or surrogate workflows.

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

Classification: **process**.

Tagged: ``dataset``, ``derivatives``, ``dydx``, ``ffd``, ``full_factorial``, ``gradient``, ``numerical_differentiation``, ``sensitivity``.

Inputs
------

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

   * - Label
     - ID
     - Type
     - Default
     - Required
     - Description
   * - Choose Dataset
     - dataset_1
     - dataset
     - —
     - 
     - Source dataset containing the input and target columns for which derivatives will be computed; accepts any tabular d3VIEW dataset.
   * - Inputs
     - inputs
     - scalar
     - —
     - 
     - One or more column names from dataset_1 to treat as the independent variables (x-axes) when computing dy/dx.
   * - Targets
     - targets
     - scalar
     - —
     - 
     - One or more column names from dataset_1 to treat as the dependent variables (y-axes) whose derivatives are computed with respect to each selected input.
   * - Build FFD If Not One
     - build_ffd
     - string
     - no
     - 
     - Whether to learn from the dataset and construct a Full Factorial Design (FFD) dataset before differentiating; set to 'yes' when the source dataset is not already a structured FFD, default 'no'.
   * - Value Type
     - value_type
     - string
     - avg
     - 
     - Strategy for aggregating derivative estimates from neighboring points: 'avg' (average of all neighbors, default), 'min', 'max', 'left' (left-side neighbor), 'right' (right-side neighbor), or 'twoside' (central difference).
   * - Normalize The Data
     - normalize
     - string
     - no
     - 
     - Whether to normalize the computed derivative values; set to 'yes' to produce dimensionless relative sensitivities, default 'no'.

Outputs
-------

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

   * - Label
     - ID
     - Type
     - Description
   * - dataset_compute_derivatives_output_1
     - dataset_compute_derivatives_output_1
     - dataset
     - Augmented tabular dataset containing the original columns plus new derivative columns named as dy_dx for each target–input pair computed.

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

- ai_ml.preprocessing
- data.dataset.transform
- design_exploration.sensitivity

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

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

.. 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_derivatives</code>. Schema hash: <code>438f5a91d717</code>. Hand-curated docs in <code>workerexamples/</code> override this page when present.
   </p>