.. _auto_assembler_evaluate_loadcase:

*ASSEMBLER EVALUATE LOADCASE*
=============================

Evaluates all simulations under one or more loadcases against their associated response templates, ranks them using a configurable method (weighted sum, threshold count, or best/worst), and identifies the top-performing simulation. Use this worker to perform structured post-simulation assessment and scoring within the d3VIEW Assembler framework.

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

Classification: **default**.

Tagged: ``assembler``, ``best_simulation``, ``evaluation``, ``kpi``, ``loadcase``, ``ranking``, ``response_template``, ``scoring``.

Inputs
------

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

   * - Label
     - ID
     - Type
     - Default
     - Required
     - Description
   * - Loadcase IDs
     - loadcase_ids
     - remote_lookup
     - —
     - ✓
     - One or more loadcase IDs whose child simulations will be evaluated; at least one is required and multiple selections are supported.
   * - Project ID
     - project_id
     - remote_lookup
     - —
     - 
     - Optional project context used to scope the loadcase lookup; automatically inferred from the selected loadcase(s) if not provided.
   * - Criteria JSON
     - criteria_json
     - json
     - —
     - 
     - Optional JSON object defining per-response evaluation thresholds and optimization direction; format: {"response_name": {"direction": "minimize"/"maximize", "threshold": <number>}}. Omit to skip criteria-based filtering.
   * - Ranking Method
     - ranking_method
     - select
     - —
     - 
     - Algorithm used to rank simulations across responses: "weighted_sum" aggregates normalized scores, "threshold_count" counts criteria met, and "best_worst" ranks by best or worst single response; defaults to "weighted_sum" when omitted.
   * - Re-extract Responses
     - reextract_responses
     - boolean
     - —
     - 
     - When true, re-triggers Simlyzer response extraction before evaluation rather than using previously stored response values; default is false (use cached values).

Outputs
-------

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

   * - Label
     - ID
     - Type
     - Description
   * - Records Dataset
     - records
     - dataset
     - Tabular dataset with one row per simulation and one column per response, providing the raw response values used for ranking.
   * - Evaluation JSON
     - evaluation_json
     - json
     - Structured JSON containing per-simulation scores, per-criteria pass/fail results, and the best simulation identified for each loadcase.
   * - Best Simulation ID
     - best_simulation_id
     - text
     - Platform ID of the top-ranked simulation (resolved from the first loadcase when multiple loadcases are evaluated).
   * - Best Simulation Name
     - best_simulation_name
     - text
     - Human-readable name of the top-ranked simulation, suitable for display or downstream reporting.
   * - Simulation Count
     - simulation_count
     - number
     - Integer count of the total number of simulations evaluated across all specified loadcases.
   * - Summary
     - summary
     - text
     - Plain-text narrative summarizing the evaluation outcome, including the best simulation and key scoring results; intended for human review or LLM agent consumption.
   * - Status
     - status
     - text
     - Short status string (e.g., "success" or an error message) indicating whether the worker completed successfully; intended for programmatic checks in agent workflows.

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

- ai_ml.agents
- cae.postprocessing.extraction
- cae.postprocessing.response
- platform.workflow

.. raw:: html

   <hr style="margin-top:2em">
   <p style="font-size:11px;color:#888">
   Auto-generated from <code>platform</code> schema. Worker id: <code>assembler_evaluate_loadcase</code>. Schema hash: <code>e35eddc6ab7f</code>. Hand-curated docs in <code>workerexamples/</code> override this page when present.
   </p>