TypeError: 'numpy.float64' object cannot be interpreted as an index about pytorch-pose HOT 4 CLOSED

same here...

from pytorch-pose.

Hi, I cannot repeat this error in my environment. Could you please provide more details about your running environment? Thank you very much!

from pytorch-pose.

Hi, I found a quick way to fix this:
Change line 141 in imutils.py to
num_cols = int(np.ceil(float(len(parts_to_show)) / num_rows))
instead of
num_cols = np.ceil(float(len(parts_to_show)) / num_rows)

the output of np.ceil is apparently a float (at least in my version).

from pytorch-pose.

@bloodymeli Thanks. The bug has been fixed.

from pytorch-pose.