Data

Data module - main class governing your data and wrapper for all other ConnectiviPy functions.

class connectivipy.data.Data(data, fs=1.0, chan_names=[], data_info='')

Class governing the communication between data array and connectivity estimators.

Args:

data : numpy.array or str

• array with data (kXNxR, k - channels nr, N - data points,

  R - nr of trials)

• str - path to file with appropieate format

fs = 1: int

sampling frequency

chan_names = []: list

names of channels

data_info = ‘’: string

other information about the data

select_channels(channels=None)

Selecting channels to plot or further analysis.

Args:

channels : list(int)

List of channel indices. If None all channels are taken into account.

filter(b, a)

Filter each channel of data using forward-backward filter filtfilt from scipy.signal.

Args:

b, a : np.array

Numerator b / denominator a polynomials of the IIR filter.

resample(fs_new)

Signal resampling to new sampling frequency new_fs using resample function from scipy.signal (basing on Fourier method).

Args:

fs_new : int

new sampling frequency

fit_mvar(p=None, method='yw')

Fitting MVAR coefficients.

Args:

p = None : int

estimation order, default None

method = ‘yw’ : str {‘yw’, ‘ns’, ‘vm’}

method of MVAR parameters estimation all avaiable methods you can find in fitting_algorithms

conn(method, **params)

Estimate connectivity pattern.

Args:

p = None : int

estimation order, default None

method : str

method of connectivity estimation all avaiable methods you can find in conn_estim_dc

short_time_conn(method, nfft=None, no=None, **params)

Short-time connectivity.

Args:

method = ‘yw’ : str {‘yw’, ‘ns’, ‘vm’}

method of estimation all avaiable methods you can find in fitting_algorithms

nfft = None : int

number of data points in window; if None, it is signal length N/5.

no = None : int

number of data points in overlap; if None, it is signal length N/10.

params

other parameters for specific estimator

significance(Nrep=100, alpha=0.05, verbose=True, **params)

Statistical significance values of connectivity estimation method.

Args:

Nrep = 100 : int

number of resamples

alpha = 0.05 : float

type I error rate (significance level)

verbose = True : bool

if True it prints dot on every realization

Returns:

signi: numpy.array

matrix in shape of (k, k) with values for each pair of channels

short_time_significance(Nrep=100, alpha=0.05, nfft=None, no=None, verbose=True, **params)

Statistical significance values of short-time version of connectivity estimation method.

Args:

Nrep = 100 : int

number of resamples

alpha = 0.05 : float

type I error rate (significance level)

nfft = None : int

number of data points in window; if None, it is taken from Data.short_time_conn() method.

no = None : int

number of data points in overlap; if None, it is taken from short_time_conn method.

verbose = True : bool

if True it prints dot on every realization

Returns:

signi: numpy.array

matrix in shape of (k, k) with values for each pair of channels

plot_data(trial=0, show=True)

Plot data in a subplot for each channel.

Args:

trial = 0 : int

if there is multichannel data it should be a number of trial you want to plot.

show = True : boolean

show the plot or not

plot_conn(name='', ylim=None, xlim=None, signi=True, show=True)

Plot connectivity estimation results.

Args:

name = ‘’ : str

title of the plot

ylim = None : list

range of y-axis values shown, e.g. [0,1] None means that default values of given estimator are taken into account

xlim = None : list [from (int), to (int)]

range of y-axis values shown, if None it is from 0 to Nyquist frequency

signi = True : boolean

if significance levels are calculated they are shown in the plot

show = True : boolean

show the plot or not

plot_short_time_conn(name='', signi=True, percmax=1.0, show=True)

Plot short-time version of estimation results.

Args:

name = ‘’ : str

title of the plot

signi = True : boolean

reset irrelevant values; it works only after short time significance calculation using short_time_significance

percmax = 1. : float (0,1)

percent of maximal value which is maximum on the color map

show = True : boolean

show the plot or not

export_trans3d(mod=0, filename='conntrans3d.dat', freq_band=[])

Export connectivity data to trans3D data file in order to make 3D arrow plots. Args:

mod = 0 : int

0 - Data.conn() results 1 - Data.short_time_conn() results

filename = ‘conn_trnas3d.dat’ : str

title of the plot

freq_band = [] : list

frequency range [from_value, to_value] in Hz.

fill_nans(values, borders)

Fill nans where values < borders (independent of frequency).

Args:

values : numpy.array

array of shape (time, freqs, channels, channels) to fill nans

borders : numpy.array

array of shape (time, channels, channels) with limes values

Returns:

values_nans : numpy.array

array of shape (time, freq, channels, channels) with nans where values were less than appropieate value from borders

mvar_coefficients

Returns mvar coefficients if calculated

mvarcoef

Returns mvar coefficients if calculated

Data loading

Additonal function which enable other data formats loading

connectivipy.load.loaders.signalml_loader(file_name)

It returns data and dictionary from SignalML files.

Args:

file_name : str

must be the same for .xml and .raw files.

Returns:

data: np.array

eeg data from raw file

xmlinfo : dict

dcitionary with keys: samplingFrequency, channelCount, firstSampleTimestamp, channelNames, calibrationCoef which means the same as in SML file

connectivipy.load.loaders.give_xml_info(path)

It returns dictionary from SignalML file.

Args:

path : str

SML file eg. ‘test.xml’

Returns:

xml_data : dict

dcitionary with keys: samplingFrequency, channelCount, firstSampleTimestamp, channelNames, calibrationCoef which means the same as in SML file

Additional functions

Plot tools which are separate from Data class

connectivipy.plot_conn(values, name='', fs=1, ylim=None, xlim=None, show=True)

Plot connectivity estimation results. Allows to plot your results without using Data class.

Args:

values : numpy.array

connectivity estimation values in shape (fq, k, k) where fq - frequency, k - number of channels

name = ‘’ : str

title of the plot

fs = 1 : int

sampling frequency

ylim = None : list

range of y-axis values shown, e.g. [0,1] None means that default values of given estimator are taken into account

xlim = None : list [from (int), to (int)]

range of y-axis values shown, if None it is from 0 to Nyquist frequency

show = True : boolean

show the plot or not