Resampled Data contains missing values about smogn HOT 5 OPEN

As a quick work around.

Go to the location, where the library is installed.

For me it was : "C:\Users\user_name\Anaconda3\envs\project_3\Lib\site-packages\smogn" and open smoter.py through notepad/notepad++ and replace the following:

    for j in range(d):
        data_new.iloc[:, j] = data_new.iloc[:, j].astype(feat_dtypes_orig[j])

with

    for j in range(d):
        data_new=data_new.fillna(data_new.median())
        data_new.iloc[:, j] = data_new.iloc[:, j].astype(feat_dtypes_orig[j])

This should fix the issue until the issue is fixed by the dev.

Note - I filled NAs with median, because it worked for me.

from smogn.

Unfortunately I could not solve the issue. I had to stay with my naive implementation of dropping columns containing nan values, which was only meant to be a backup solution.

from smogn.

@mplutat Hello and thank you for raising this issue. Is there some you could provide to reproduce the problem? This sounds like it might be an issue with the DataFrame. However, I cannot tell from this alone.

Also, I would not be able to responsibly comment on the affect of simply removing missing values, as I have not seen the distribution / change in distribution.

from smogn.

I'm sorry for the delay. It took longer than expected to find and export affected DataFrames.
Anyway I indentified one bin which when over-sampled contains missing values in the synthetic DataFrame.
I have attached the input and resampled bin DataFrames to this comment. In case the given DataFrames are not thorough enough I have also attached the complete input DataFrame containing all values.
Thank you very much for your help.
dataframes.zip

from smogn.

Managed to solve?
I have the same problem.

from smogn.