statsmodels.stats.proportion.proportions_chisquare_allpairs has hardcoded value #1493

mtrbean opened this Issue Mar 19, 2014 · 1 comment


None yet

2 participants

mtrbean commented Mar 19, 2014

This should be a quick fix:

all_pairs = zip(*np.triu_indices(4, 1))

should be

all_pairs = zip(*np.triu_indices(len(count), 1))
@josef-pkt josef-pkt added this to the 0.5.1 milestone Mar 19, 2014

Thanks @mtrbean

clearly a bug, there is only one test case and that has 4 values.

aside: the smprop.proportions_chisquare_pairscontrol test case doesn't check that the multiple testing p-value correction has the correct values, only raw p-values have a test.

proportions_chisquare_allpairs also has verified numbers for 'holm' corrected p-values

@josef-pkt josef-pkt closed this in #1613 Apr 21, 2014
@PierreBdR PierreBdR pushed a commit to PierreBdR/statsmodels that referenced this issue Sep 2, 2014
@josef-pkt josef-pkt BUG: proportions_chisquare_allpairs, fix hardcode pairs closes #1493 d5f5540
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment