cf9612933e79ba8ed6457b5013f9aa9f8f3376f5,nisl/signals.py,,high_variance_confounds,#Any#Any#Any#,174

Before Change

        // Adding a transposition gives F order computation.
        var = np.mean((series.T ** 2).T, axis=0)
    else:
        var = np.mean(series ** 2, axis=0)

    var_thr = stats.scoreatpercentile(var, 100. - percentile)
    series = series[:, var > var_thr]  // extract columns (i.e. features)
    // Return the singular vectors with largest singular values

After Change

    // Compute variance without mean removal.
    // The execution speed of these three lines is independent of array
    // ordering (C or F)
    var = np.copy(series)
    var **= 2
    var = var.mean(axis=0)

    var_thr = stats.scoreatpercentile(var, 100. - percentile)

Instances