# ---
# jupyter:
#   jupytext:
#     text_representation:
#       extension: .py
#       format_name: sphinx
#       format_version: '1.1'
#       jupytext_version: 1.10.2
#   kernelspec:
#     display_name: Python 3
#     language: python
#     name: python3
# ---

"""
.. _exa-utsstat:

UTS-statistics (``UTSStat``)
============================

"""
# sphinx_gallery_thumbnail_number = 5
from eelbrain import *


ds = datasets.simulate_erp(snr=1)
# add an orthogonal variable
ds['length'] = Factor(ds['n_chars'] > 4, labels={True: 'long', False: 'short'})
# smooth the data to make plots nicer
ds['cz'] = ds["eeg"].sub(sensor='Cz').smooth('time', 0.05)

###############################################################################
# Plot the mean and SEM by condition:

p = plot.UTSStat('cz', 'cloze_cat', ds=ds, axh=3)

###############################################################################
# Split between different axes and use a ``t``-frame:

p = plot.UTSStat('cz', 'cloze_cat', 'length', ds=ds, frame='t', axh=2)

###############################################################################
# Colors can also be set using the within-plot category only:

colors = plot.colors_for_oneway(['high', 'low'])
p = plot.UTSStat("cz", 'cloze_cat', 'length', ds=ds, colors=colors, frame='t', legend=False, axh=2)
legend = plot.ColorList(colors)

###############################################################################
# Plot all categories in the same plot. Make lines more distinguishable through line-style:

colors = {
    ('high', 'long'): plot.Style((1, 0, 0)),
    ('high', 'short'): plot.Style((1, 0.5, 0.5), linestyle='--'),
    ('low', 'long'): plot.Style((0, 0, 1)),
    ('low', 'short'): plot.Style((0.5, 0.5, 1), linestyle='--'),
}

p = plot.UTSStat("cz", 'cloze_cat % length', ds=ds, colors=colors, frame='t', top=1e-6, axh=3)