lenskit.tuning.PipelineTuner
============================

.. py:class:: lenskit.tuning.PipelineTuner(spec, out_dir = None, rng = None)
   :canonical: lenskit.tuning._search.PipelineTuner

   Set up and run a hyperparameter tuning job for a pipeline.

   :Stability: Experimental


   .. py:attribute:: settings
      :type:  lenskit.config.TuneSettings


   .. py:attribute:: spec
      :type:  lenskit.tuning.spec.TuningSpec


   .. py:attribute:: out_dir
      :type:  pathlib.Path


   .. py:attribute:: pipe_name
      :type:  str | None


   .. py:attribute:: random_seed
      :type:  numpy.random.SeedSequence


   .. py:attribute:: iterative
      :type:  bool


   .. py:attribute:: data
      :type:  lenskit.splitting.TTSplit


   .. py:attribute:: harness
      :type:  Any


   .. py:attribute:: tuner
      :type:  ray.tune.Tuner

      The Ray tuner that is used for tuning.  Not available until :meth:`setup`
      has been called.


   .. py:attribute:: results
      :type:  ray.tune.ResultGrid

      Ray tuning results. Only available after :meth:`run` has been called.


   .. py:attribute:: log


   .. py:property:: mode


   .. py:method:: set_data(train, test, *, name = None)

      Set the data to be used for tuning.



   .. py:method:: setup()

      Set up to run the trainer.  After this method completes, the
      :attr:`tuner` is ready.



   .. py:method:: run()

      Run the tuning job.

      Saves the results in :attr:`results`, and also returns them.



   .. py:method:: best_result(*, scope = 'all')

      Get the best configuration and its validation metrics.

      :param scope: The metric search scope for iterative training.  Set to
                    ``"last"`` to use the last iteration instead of the best
                    iteration.  See :meth:`ray.tune.ResultGrid.get_best_result` for
                    details.



   .. py:method:: best_pipeline()

      Get the (full) configuration for the best pipeline.



   .. py:method:: search_space()

      Get the Ray search space.



   .. py:method:: setup_harness()


   .. py:property:: metric


   .. py:method:: create_tuner()

      Create a Ray tuner for the search.