filter.py

import subprocess

import numpy as np
from scipy import sparse, io

threshold = 1000
folders = ['M7_5', 'M8_3', 'X5.3.4', 'X5_2', 'X6.1']

for f in folders:
    print(f)
    data = sparse.csc_matrix(io.mmread(f + '/matrix.mtx.gz'))
    print('data loaded')
    cell_sums = np.array(data.sum(0)).flatten()
    data_filtered = data[:, cell_sums > threshold]
    print('data filter calculated')
    # TODO: save indices?
    np.savetxt(f + '/filtered_indices.txt', cell_sums > threshold, fmt='%d')
    data_path_new = f + '/matrix_filtered.mtx'
    io.mmwrite(data_path_new, data_filtered)
    print('filtered data saved')
    subprocess.call(['gzip', data_path_new])