"""
===========================
Neo All - example 2 - Cross-Correlograms
===========================

This example shows how to compute and plot cross-correlograms of spiketrains from different units.

.. note:: The crosscorrelogram compares the output of 2 different neurons, it indicates the firing rate of one neuron
          versus another. See `here <https://www.med.upenn.edu/mulab/crosscorrelation.html>`_ for more details

.. image:: ./../../_static/images/crosscorrelogram.png


"""

############################
# First import neoStructures
from neoStructures import *
import matplotlib.pyplot as plt
from os.path import isdir, join

#######################
# Import the data and create the NeoAll instance
data_dir = join('pySpikeAnalysis', 'sample_data') if isdir('pySpikeAnalysis') else join('..', '..', 'pySpikeAnalysis', 'sample_data')
spykingcircus_dir = r'SpykingCircus_results'
probe_filename = r'000_AA.prb'
results_filename = r'spykingcircusres'

neoAll = NeoAll(join(data_dir, spykingcircus_dir), results_filename, join(data_dir, probe_filename), save_fig=0)

##############################
# See information about NeoAll
print(neoAll)

##############################
# Use :func:`neoStructures.NeoAll.plot_crosscorrelogram` to plot cross-correlogram between 2 units. The spiketrains are first converted into binned
# spiketrains before the computation of the cross-correlogram.
# The package `Elephant <http://elephant.readthedocs.io/en/latest/index.html>`_ is used for the binning as well as for computing
# the cross-correlogram.
# Let's compute the cross-correlogram between the first 2 units :
neoAll.plot_crosscorrelogram(0, 1)

##############################
# We can see from these cross-correlogram that the two units often fire together
# Bin duration is set by default to 1ms but can be modified. The max_lag_time parameter sets the time limits of the
# cross-correlogram, its default value is set to 80 ms
# It can be changed to zoom on the peak near the origin :
neoAll.plot_crosscorrelogram(0, 1, bin_time=1*ms, max_lag_time=25*ms)

##############################
# Some statistics can be computed, be setting the do_stat parameter to 1 : n_surrogates spike-trains are created
# in which a jitter is added to the time of the spikes. The jitter is computed from a normal distribution whose standard
# deviation is fixed by the normal_dist_sd parameter. The 99% confidence interval computed from the jittered spiketrains
# is shown on top of the cross-correlogram.
neoAll.plot_crosscorrelogram(0, 1, do_stat=True, n_surrogates=20, normal_dist_sd=25*ms)

##############################
# If unit_pos_a and unit_pos_b parameters are equals, the autocorrelogram is computed.
neoAll.plot_crosscorrelogram(0, 0)

##############################
# Multiples cross-correlogram can be plot at the same time in multiple figures :
neoAll.plot_crosscorrelogram(0, [0, 1, 2])

##############################
# Or in the same figure :
neoAll.plot_crosscorrelogram([0, 1, 2], [0, 1, 2], merge_plots=1)

##############################
# If same_yscale is True, the cross-correlograms are smoothed and the same y-scale is used.
neoAll.plot_crosscorrelogram([0, 1, 2], [0, 1, 2], merge_plots=1, same_yscale=1, fill_under_plot=1)