Conpute SVD

import matplotlib.pyplot as plt
import numpy as np

import pycartool as cart

#### Import from Cartool files

fname_spi = "MNI152.NlinAsym09c.204.5000.2017.spi"
fname_roi = "MNI152.NlinAsym09c.204.5000.2017.AAL.rois"

spi = cart.spi.read_spi(fname_spi)
roi = cart.rois.read_roi(fname_roi, spi)

#### Constants

sfreq = 512
n_sources = len(spi.names)
snr = 10
n_times = 2048

### Simulate sources time course

Here we simulate the sources time course as random noise generators everywhere in the brain, expect in the first Region of interest where sources time course are simulated are sin waves in the x direction.

#### Create random noise

simulated_tc = np.random.normal(size=(n_sources, 3, n_times))

plt.figure()
plt.plot(simulated_tc[0, 0], color="navy")
plt.title("Time course of a source outside the first Roi")
plt.show()

#### Create sin wave in Roi

x = np.arange(0, n_times, 1)
sin = snr * np.sin(x / (2 * np.pi))

for elem in roi.groups_of_indexes[0]:
    simulated_tc[elem][0] = sin
    simulated_tc[elem][1] = np.zeros(sin.shape)
    simulated_tc[elem][2] = np.zeros(sin.shape)

plt.figure()
plt.plot(sin, color="red")
plt.title("Time course of sources inside the first Roi  (x direction)")
plt.show()

source_estimate_simulated = cart.ris.SourceEstimate(
    simulated_tc, sfreq=sfreq, source_space=spi
)

#### Compute the regions of interest time course

roi_t_simulated = source_estimate_simulated.compute_rois_tc(roi)

plt.figure()
plt.plot(roi_t_simulated.sources_tc[0:4, 0, :].T)
plt.title("Rois time course")
plt.show()

Estimated memory usage: 502 MB