_bar_df wrong for loop boundary

[MaaikeG]

In PyEMMA/pyemma/thermo/extensions/bar/_bar.c

The second for loop should run from 0 to L2, not L1. Current implementation gives an access violation whenever L1 > L2

extern double _bar_df(double *db_IJ, int L1, double *db_JI, int L2, double *scratch)
{
    int i;
    double ln_avg1;
    double ln_avg2; 
    for (i=0; i<L1; i++)
    {
        scratch[i] = db_IJ[i]>0 ? 0 : db_IJ[i];
    }
    ln_avg1 = _logsumexp_sort_kahan_inplace(scratch, L1);
    for (i=0; i<L1; i++)
    {
        scratch[i] = db_JI[i]>0 ? 0 : db_JI[i];
    }
    ln_avg2 = _logsumexp_sort_kahan_inplace(scratch, L2);
    return ln_avg2 - ln_avg1;
}```

[clonker]

Ouch :) To be fair I think we should just wait until deeptime-ml/deeptime#168 is merged and then back-integrate here with the next deeptime release.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.

[stale]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.