ivneuro.delta_power_spectral

A module with delta_power_spectral function.

Functions

delta_power_spectral(contvar, exp_interval, baseline)

Calculate the power spectral at a experimental time of interest (exp_interval), at a baseline interval (baseline), and the frequency-basis normalized difference betrween the experimental interval and the baseline.

Module Contents

ivneuro.delta_power_spectral.delta_power_spectral(contvar, exp_interval, baseline, lowest_freq=0, highest_freq=500, sample_subsamples=None, sampling_rate=None, nfft=2000, scaling='spectrum')

Calculate the power spectral at a experimental time of interest (exp_interval), at a baseline interval (baseline), and the frequency-basis normalized difference betrween the experimental interval and the baseline.

This function is based on the periodogram function of scipy.signal. exp_interval and baseline must be lists of slices the same lenght, when when the lists have more than one element each, the results are the averages across intervals (either power spectral at exp_interval, power spectral at baseline, or their normalized difference). To normalize, it uses the formula: (Power(exp_interval) - Power(baseline)) / (Power(exp_interval) + Power(baseline)) for each frequency and trial (pair of elements in interval and baseline), where Power(exp_interval) is the power at the experimental interval and Power(baseline) is the power at the baseline.

Parameters:

  • contvar (pandas.DataFrame) – Dataframe with continuous variables in each column, and timestamps as index.

  • exp_interval (list) – List of slices corresponding to experimental intervals to calculate power.

  • baseline (list) – List of slices corresponding to baseline intervals to calculate power.

  • lowest_freq (int or float, optional) – The lowest frequency to include in the power spectral. The default is 0.

  • highest_freq (int or float, optional) – The highest freq to include in the power spectral. The default is 500.

  • sample_subsamples (dict or None, optional) – When a dict, key must be sample names and values must be lists of subsample names, e.g.: {‘sample_A’:[‘subsample_A1’,’subsample_A2’…’subsample_An’], ‘sample_B’:[‘subsample_B1’,’subsample_B2’…’subsample_Bn’]…}. Power spectrals and normalized delta power spectral will be averaged across subsamples of the same sample. It must only be used when contvar contains multiple replicates for each observation (e.g., local field potentials recorded with multiple wires of the same tetrode). If None, each column of contvar is treated as an independent sample, and therefore the results contain spectrals of every contvar column. The default is None.

  • sampling_rate (int or float or None, optional) – Sampling rate of the continuous variables. If None, the sampling_rate is calculated using the inverse of the median difference between consecutive timestamps of the contvar’s index. The default is None.

  • nfft (int, optional) – The nfft parameter to enter to signal.spectrogram() function of the scipy package. Refer to scipy.signal.spectogram in scipy manual for more information. The default is 2000.

  • scaling (str, optional) – The scaling parameter to enter to signal.spectrogram() function of the scipy package. Refer to scipy.signal.spectogram in scipy manual for more information. The default is ‘spectrum’.

Returns:

  • power_exp_interval (pandas.DataFrame) – Power spectral for each variable (or sample if sample_subsamples != None), at the experimental interval. Variable names (or sample names) as columns and frequency as index.

  • power_baseline (pandas.DataFrame) – Power spectral for each variable (or sample if sample_subsamples != None), at the baseline interval. Variable names (or sample names) as columns and frequency as index.

  • delta_power (pandas.DataFrame) – Normalized difference of power spectrals between experimental interval and baseline interval, for each variable (or sample if sample_subsamples != None). Variable names (or sample names) as columns and frequency as index.

Examples

Create event, intervals and signals.

>>> import ivneuro as ivn
>>> import pandas as pd
>>> event = [*range(30,300, 30)] # Events
>>> # Make intervals
>>> exp_interval = ivn.make_intervals(event, 0, 2) # Experimental interval
>>> baseline = ivn.make_intervals(event, -6, -4) #Baseline interval
>>> # Create signals
>>> signal1 = ivn.generate_signal(300, event, 30, burst_amplitude=0.06, seed=15)
>>> signal2 = ivn.generate_signal(300, event, 30.2, burst_amplitude=0.13, seed = 30)
>>> signal3 = ivn.generate_signal(300, event, 80, burst_amplitude=0.05, seed = 50)
>>> signals = pd.concat([signal1, signal2, signal3], axis = 1)

Calculate power spectral and delta power spectral between intervals with delta_power_spectral function.

>>> exp_ps, baseline_ps, delta_ps = ivn.delta_power_spectral(signals, exp_interval, baseline)
>>> delta_ps.head()
     Signal 30Hz  Signal 30.2Hz  Signal 80Hz
0.0    -0.110932      -0.112460    -0.006010
0.5    -0.087769       0.009546     0.149461
1.0     0.180706       0.042875     0.142876
1.5    -0.167544      -0.215748     0.257452
2.0     0.108177      -0.357427     0.086913

Set signal1 and signal2 as replicates of sample1, and signal3 as sample2.

>>> # Make dictionary to group replicates of each sample
>>> sample_subsamples = {'sample1' : ['Signal 30Hz', 'Signal 30.2Hz'], 'sample2':['Signal 80Hz']}
>>> # Calculate coherence and delta coherence
>>> exp_ps, baseline_ps, delta_ps = ivn.delta_power_spectral(signals, exp_interval, baseline, sample_subsamples = sample_subsamples)
>>> delta_ps.head()
      sample1   sample2
0.0 -0.111696 -0.006010
0.5 -0.039112  0.149461
1.0  0.111790  0.142876
1.5 -0.191646  0.257452
2.0 -0.124625  0.086913