Python z-score replace by mean or some other value

Asked Apr 12 '20 at 12:31

Active Apr 12 '20 at 13:13

Viewed 291 times

I have sensor data in the form of a pandas DataFrame along three columns (X,Y,Z). I thought about removing the rows where a row has in any column a z-score higher than 2.5 by doing

df = df[(np.abs(stats.zscore(df)) < 2.5).all(axis=1)]

But this discards the entire row so also for other columns which may have normal values. What should be the proper way to replace these abnormal high (or low) values by the mean or maybe a value closer to the mean?

edited Apr 12 '20 at 13:13

asked Apr 12 '20 at 12:31

Wouter Vandenputte

1,948
4
26
50

Look at this https://stackoverflow.com/questions/13851535/delete-rows-from-a-pandas-dataframe-based-on-a-conditional-expression-involving – Sheri Apr 12 '20 at 12:35

Python z-score replace by mean or some other value

0 Answers0