Eelbrain

-----------

* API:

- :class:`plot.Correlation` renamed to :class:`plot.Scatter` with some parameter changes for improved functionality.

* New:

- :func:`boosting`: Option to store TRFs for the different test partitions
(``partition_results`` parameter).
- :func:`normalize_in_cells` (see :ref:`exa-compare-topographies`).
- :class:`UTS` dimension: ``unit`` parameter to represent time in units other than seconds.
- :mod:`report` submodule with shortcuts for data summary and visualization.
- :func:`load.convert_pickle_protocol` for compatibility with older Python version.

-----------

* API :func:`load.mne.events`: The merge parameter is now by default inferred based on the raw data.
* Boosting: plot data partitioning scheme (``BoostingResult.splits.plot()``).
* :class:`~pipeline.MneExperiment` pipeline:

- New :attr:`pipeline.MneExperiment.merge_triggers` attribute.
- Compatibility with Microsoft Windows.

-----------

.. currentmodule:: eelbrain

* Requires at least `Python 3.7 <https://docs.python.org/3.7/>`_
* API changes:

- Consistent class names for tests in :mod:`test`, :mod:`testnd` and :mod:`pipeline`.
- :class:`plot.Timeplot` argument order: second and third argument switched to facilitate plotting single category.
- Significance markers for trends (.05 < *p* ≤ .1) are disabled by default.

* :func:`boosting`:

- When using a ``basis``, the function now considers the effect of the basis when normalizing predictors. This change leads to slightly different results, so TRFs should not be compared between this and previous versions. To facilitate keeping track of this change, it corresponds to a change in the :attr:`BoostingResult.algorithm_version` attribute from ``-1`` to ``0``.
- Different ``tstart``/``tstop`` for different predictors (contributed by `Joshua Kulasingham`_)
- Cross-validation of model fit (``test`` parameter)

* :class:`plot.Style` to control advanced plotting options by category (see :ref:`exa-boxplot` example).
* New functions/methods:

- :meth:`NDVar.quantile`

* :class:`~pipeline.MneExperiment` pipeline:

- Methods with ``decim`` parameter now also have ``samplingrate`` parameter
- More control over :ref:`MneExperiment-events`

-----------

* API changes:

- :class:`Var` and :class:`NDVar` argument order changed to be consistent with other data objects
- :func:`combine`: Combining :class:`NDVar` with unequal dimensions will now raise an error; to combine them by taking the intersection of valid elements (previous behavior), use ``dim_intersection=True``
- :meth:`Dataset.save_txt`: ``delim`` parameter renamed to ``delimiter``
- :mod:`testnd` API: For permutation tests, the ``samples`` parameter now defaults to 10,000 (previously 0)
- :func:`table.difference`: The ``by`` parameter is deprecated, use ``match`` instead
- :meth:`NDVar.smooth` with a window with an even number of samples, and :attr:`BoostingResult.h` for :func:`boosting` with a basis with an even number of samples: the time axis is now consistent with :func:`scipy.signal.convolve` (was previously shifted by half a sample)
- :meth:`testnd.LMGroup.coefficients_dataset` now returns a wide form table by default
- :meth:`plot.Topomap.mark_sensors`, :meth:`plot.TopomapBins.mark_sensors` and :meth:`plot.SensorMap.mark_sensors`: The second argument now specifies axis to mark

* New functions:

- :func:`gaussian`
- :func:`powerlaw_noise`
- :func:`set_time`
- :func:`plot.two_step_colormap`

* :class:`plot.Boxplot`: Accepts additional arguments (``label_fliers`` and :meth:`~matplotlib.axes.Axes.boxplot` parameters)
* :class:`plot.BarplotHorizontal`: Horizontal bar-plot
* Non-parametric univariate tests :class:`test.MannWhitneyU` and :class:`test.WilcoxonSignedRank`
* :class:`~pipeline.MneExperiment` pipeline:

- :class:`pipeline.SubParc`: Simplified subset parcellation

-----------

* Support for vector data (with many contributions from `Proloy Das`_):

- :class:`Space` dimension to represent physical space
- :class:`VolumeSourceSpace` to represent volume source spaces
- Statistical tests: :class:`testnd.Vector`, :class:`testnd.VectorDifferenceRelated`
- Plotting with :class:`plot.GlassBrain`

* ICA-GUI: tool to find high amplitude signals
* Documentation: New :ref:`examples` section
* :meth:`Dataset.summary` method
* Element-wise :func:`maximum` and :func:`minimum` functions for :class:`NDVar` objects
* :class:`~pipeline.MneExperiment` pipeline:

- :class:`~pipeline.RawApplyICA` preprocessing pipe to apply ICA estimated in a different branch of the pipeline.
- :meth:`pipeline.MneExperiment.load_evoked_stc` API more closely matches :meth:`pipeline.MneExperiment.load_epochs_stc`
- :meth:`pipeline.MneExperiment.load_neighbor_correlation`

-----------

* API: Better default parameters for :func:`resample`
* Predictor-specific stopping for :func:`boosting`
* New :class:`NDVar` function :func:`correlation_coefficient`
* Plotting:

- :ref:`general-layout-parameters` for plot size relative to screen size
- Better plots for masked statistic maps

* :class:`~pipeline.MneExperiment` pipeline:

- API: :meth:`pipeline.MneExperiment.make_rej` renamed to :meth:`pipeline.MneExperiment.make_epoch_selection`
- Object-based definitions (see :ref:`experiment-class-guide`)
- New method :meth:`pipeline.MneExperiment.plot_raw`