.. _fwflag_factor_adaptation:
=====================
--factor_adaptation
=====================
Switch
======

--factor_adaptation

Description
===========

Applies factor adaptation: rather than using the outcome controls directly, selects a smaller number of derived "factors" from them (or from :doc:`fwflag_adaptation_factors`) and adapts language features against those. Aliased as ``--fa``.

Argument and Default Value
==========================

No argument. Boolean switch, default: ``False``. Only takes effect with :doc:`fwflag_combo_test_regression`.

Details
=======

Verified end-to-end with :doc:`fwflag_factor_selection_type` ``pca`` (the default ``rfe`` is currently broken — see the Known issue on that page). If :doc:`fwflag_adaptation_factors` isn't given, falls back to using all of :doc:`fwflag_outcome_controls` as the factor pool. Sets an internal ``integrationMethod`` of ``"fa"`` (or ``"fa_plus"`` if combined with :doc:`fwflag_factor_addition`).

Other Switches
==============

Required Switches:

* :doc:`fwflag_combo_test_regression`
* :doc:`fwflag_outcome_controls`

Optional Switches:

* :doc:`fwflag_adaptation_factors`, :doc:`fwflag_num_of_factors`, :doc:`fwflag_factor_selection_type`
* :doc:`fwflag_paired_factors`, :doc:`fwflag_factor_addition`, :doc:`fwflag_report`

Example Commands
================

Using the packaged ``dla_tutorial`` dataset (see :doc:`../../tutorials/tut_dla`):

.. code-block:: bash

	dlatkInterface.py -d dla_tutorial -t msgs -c user_id -f 'feat$1to3gram$msgs$user_id$0_05' \
	--outcome_table blog_outcomes --outcomes age --outcome_controls gender is_student is_education \
	--combo_test_regression --factor_adaptation --adaptation_factors gender is_student is_education \
	--num_of_factors 2 --factor_selection_type pca