Question about eval university1652 code

[szyfh]

Great work!
In function compute_mAP
`def compute_mAP(index, good_index, junk_index):
ap = 0
cmc = torch.IntTensor(len(index)).zero_()
if good_index.size==0: # if empty
cmc[0] = -1
return ap,cmc

# remove junk_index
mask = np.in1d(index, junk_index, invert=True)
index = index[mask]

# find good_index index
ngood = len(good_index)
mask = np.in1d(index, good_index)
rows_good = np.argwhere(mask==True)  
rows_good = rows_good.flatten()

cmc[rows_good[0]:] = 1
for i in range(ngood):
    d_recall = 1.0/ngood
    precision = (i+1)*1.0/(rows_good[i]+1)
    if rows_good[i]!=0:
        old_precision = i*1.0/rows_good[i]
    else:
        old_precision=1.0
    ap = ap + d_recall*(old_precision + precision)/2

return ap, cmc`

rows_good represents the indices after filtering out junk_index, while cmc is of the original length. This causes a mismatch in indices. So is there any code mistakes？
Looking forward to your reply.