BUG: Fix unwanted type casting while replacing values in a DataFrame #33067
Conversation
phofl
changed the title
|
Thank, I'll follow this way in the future. I added a test. When you use a value, which is actually replaced in the test, the pd.NA is converted to np.nan, if the original DataFrame had float64 as datatype. The example given in the issue did not replace a value from the original DataFrame. Don't know, if this is a desired behavior. |
| to_replace=to_replace, | ||
| value=value, | ||
| inplace=inplace, | ||
| filter=filter, | ||
| regex=regex, | ||
| convert=convert, | ||
| ) | ||
| blocks_converted = [] | ||
| for ls_elem in block_replaced: |
There was a problem hiding this comment.
why is this logic not just in Block.replace itself? seems too complicated here
There was a problem hiding this comment.
I wanted to ensure, that the conversion would only take place, after the block was forcefully cast to an ObjectBlock before.
I can add it to the replace function, if that would be better.
There was a problem hiding this comment.
I wanted to ensure, that the conversion would only take place, after the block was forcefully cast to an ObjectBlock before.
I can add it to the replace function, if that would be better.
the regression was caused by a clean-up to Block.replace (#27768), see #32988 (comment)
so fixing Block.replace as @jreback suggests seems more appropriate
simonjayhawkins
added
the
Regression
|
i've labelled this as a regression, see #32988 (comment) cc @jbrockmendel |
|
can you merge master and see if you can simplify |
| to_replace=to_replace, | ||
| value=value, | ||
| inplace=inplace, | ||
| filter=filter, | ||
| regex=regex, | ||
| convert=convert, | ||
| ) | ||
| blocks_converted = [] | ||
| for ls_elem in block_replaced: |
There was a problem hiding this comment.
I wanted to ensure, that the conversion would only take place, after the block was forcefully cast to an ObjectBlock before.
I can add it to the replace function, if that would be better.
the regression was caused by a clean-up to Block.replace (#27768), see #32988 (comment)
so fixing Block.replace as @jreback suggests seems more appropriate
replacecasts columns toobject#32988black pandasgit diff upstream/master -u -- "*.py" | flake8 --diffAs mentioned in #32988 here I think, that I have found a way to fix this, but I don't know, if this results in a desired behavior.
I would add tests after ensuring, that this does not break anything else.
Any thoughts about this?