The indexes of DataFrame.describe(percentiles=[0.29, 0.57, 0.58]) are incorrect #48298
Conversation
mroeschke
added
the
Algos
pandas/io/formats/format.py
Outdated
| @@ -1721,10 +1721,10 @@ def format_percentiles( | |||
|
|
|||
| percentiles = 100 * percentiles | |||
|
|
|||
| int_idx = np.isclose(percentiles.astype(int), percentiles) | |||
| int_idx = np.isclose(percentiles.round().astype(int), percentiles) | |||
There was a problem hiding this comment.
Could you make percentiles.round().astype(int) a dedicated variable so it can be used here and below?
Co-authored-by: Matthew Roeschke <[email protected]>
|
|
||
| if np.all(int_idx): | ||
| out = percentiles.astype(int).astype(str) | ||
| out = percentiles_round_type.astype(str) |
There was a problem hiding this comment.
Does the example in the original issue hit this path or the one below? Just want to make sure we have 2 tests that didn't work before and hit this change and the one below to work correctly now
There was a problem hiding this comment.
Does the example in the original issue hit this path or the one below? Just want to make sure we have 2 tests that didn't work before and hit this change and the one below to work correctly now
a) The example from issue will hit this path.
pd.DataFrame(data=[1, 2, 3], columns=['value']).describe(percentiles=[0.29, 0.57, 0.58])
if np.all(int_idx):
out = percentiles_round_type.astype(str)
return [i + "%" for i in out]
b) The example below I own create will hit below path.
pd.DataFrame(data=[1, 2, 3], columns=['value']).describe(percentiles=[0.291, 0.57, 0.58])
out[int_idx] = percentiles[int_idx].round().astype(int).astype(str)
There was a problem hiding this comment.
Could you add the pd.DataFrame(data=[1, 2, 3], columns=['value']).describe(percentiles=[0.29, 0.57, 0.58]) case as a unit test?
|
Thanks @weikhor! |
… incorrect (pandas-dev#48298) * add test * add test * add * Update doc/source/whatsnew/v1.6.0.rst Co-authored-by: Matthew Roeschke <[email protected]> * add Co-authored-by: Matthew Roeschke <[email protected]>
DataFrame.describe(percentiles=[0.29, 0.57, 0.58])are incorrect #46362doc/source/whatsnew/vX.X.X.rstfile if fixing a bug or adding a new feature.