MODEL_METRICS
=============

Python object: ``phenotypic.schema._model_metrics.MODEL_METRICS``

Generic fit-quality metrics and diagnostics shared by all ModelFitter subclasses.

These columns are produced by any model fitter that wraps
``scipy.optimize.least_squares``, independent of the specific
mathematical model. Subclass-specific fitted parameters live in the
subclass's own MeasurementInfo class (e.g., ``LOG_GROWTH_MODEL``).

.. list-table:: Category: **ModelMetrics**
   :header-rows: 1

   * - Column label
     - Description
   * - ``ModelMetrics_MAE``
     - The mean absolute error
   * - ``ModelMetrics_MSE``
     - The mean squared error
   * - ``ModelMetrics_RMSE``
     - The root mean squared error
   * - ``ModelMetrics_R2``
     - The coefficient of determination
   * - ``ModelMetrics_NumSamples``
     - The number of samples used for model fitting
   * - ``ModelMetrics_OptimizerLoss``
     - The loss of model fitting
   * - ``ModelMetrics_OptimizerStatus``
     - The output of the optimizer status