clophfit.fitting.pipeline
=========================

.. py:module:: clophfit.fitting.pipeline

.. autoapi-nested-parse::

   Pipeline orchestrators for fitting multistage workflows (e.g., FGLS).



Functions
---------

.. autoapisummary::

   clophfit.fitting.pipeline.fgls_plate_fit
   clophfit.fitting.pipeline.fit_plate


Module Contents
---------------

.. py:function:: fgls_plate_fit(datasets, sigma_floor, *, first_pass_method = 'huber', second_pass_method = 'lm')

   Two-stage Feasible Generalized Least Squares (FGLS) plate fit.

   1. First-pass fit (typically robust like 'huber') on each well.
   2. Extract residuals globally and calibrate the comprehensive error model,
      anchoring the constant noise term to the provided ``sigma_floor``.
   3. Second-pass fit using the exact pooled weights derived from the model.

   :param datasets: The dataset dictionary keyed by well name.
   :type datasets: dict[str, Dataset]
   :param sigma_floor: Known read-noise floor per label (e.g. from buffer wells).
   :type sigma_floor: dict[str, float]
   :param first_pass_method: Method for the first-pass fit (default 'huber').
   :type first_pass_method: str
   :param second_pass_method: Method for the second-pass, calibrated fit (default 'lm').
   :type second_pass_method: str

   :returns: Final fit results and the calibrated error model parameters
             (sigma_read, gain, alpha) for each label.
   :rtype: tuple[dict[str, FitResult[typing.Any]], dict[str, tuple[float, float, float]]]


.. py:function:: fit_plate(datasets, method = 'lm', **kwargs)

   Run a single-pass fit on an entire plate of datasets.

   :param datasets: A mapping of well keys (e.g. 'A01') to `Dataset` objects.
   :type datasets: dict[str, Dataset]
   :param method: The fitting method to use: 'lm' (default), 'odr', or 'mcmc'.
   :type method: str
   :param \*\*kwargs: Additional keyword arguments passed to the specific fitting function.
   :type \*\*kwargs: typing.Any

   :returns: A dictionary mapping well keys to their corresponding `FitResult`.
   :rtype: dict[str, FitResult[typing.Any]]