Delete specific strings from pandas dataframe with operators chaining

Question

I want to delete specific strings with regular expressions from the column Sorte which I don't want to have in my dataframe file_df with the following code:

file_df = file_df[(file_df.Sorte != 'sonstige') & (file_df.Sorte != 'verauslagte Portokosten')
                  & (file_df.Sorte != 'erhaltenenzahlung Re  vom')
                  & (file_df.Sorte != 'geleistetenzahlung aus Re-Nr')
                  & (file_df.Sorte != '^.*Holzkisten geliefert.*$')
                  & (file_df.Sorte != '^.*Infomaterialktionspakete.*$')
                  & (file_df.Sorte != '^.*Aloe Vera  haben wir nicht im Sortiment.*$') 
                  & (file_df.Sorte != '^.*Anzeigenvorlage Planten ut`norden.*$')]

But somehow when I execute this code these strings still are in the dataset and I can not figure out why. I wanted to chain this expression to not create so many copies.

Update

The code worked for some strings in the dataset, for others not.

You are **not** using regular expressions here, but simple string comparison — mozway, Oct 12 '21 at 07:51
Should you be using OR '|' instead ? You code seems to be saying Var != 'A' & Var != 'B'. Can't be A and B at the same time. — EBDS, Oct 12 '21 at 08:18
@mozway How can I compare strings using regular expressions? — jeiglsperger, Oct 12 '21 at 08:22
@EBDS As all the strings are in the column "Sorte", I guess that "&" should work as the condition that all the unwanted strings shouldn't be in the remaining dataset can be fulfilled, but I can try "|" — jeiglsperger, Oct 12 '21 at 08:29
I think could try == and "|" since you want to remove them. Then use ~ to negate it. — EBDS, Oct 12 '21 at 08:58

score 0 · Answer 1 · answered Oct 12 '21 at 08:07

0

Maybe something like this:

ls = ['sonstige', 'verauslagte Portokosten', 'erhaltenenzahlung Re  vom', ...]
file_df = file_df[~ file_df.country.str.contains('|'.join(ls))]

answered Oct 12 '21 at 08:07

tzinie

717
5
9

jeiglsperger · Accepted Answer · 2022-08-08T10:04:45.187

I figured out another solution that worked out for me derived from the answer at https://stackoverflow.com/a/54410702/14553595, which is basically a combination of your suggestions in the comments:

file_df = file_df.loc[:,~(file_df.columns.str.contains('^.*Fracht.*$', case=False)
                          | file_df.columns.str.contains('^.*Angebotspaket.*$', case=False)
                          | file_df.columns.str.contains('^.*Werbe.*$', case=False)
                          | file_df.columns.str.contains('^.*Vita.Verde.*$', case=False)
                          | file_df.columns.str.contains('^.*zahlung.*$', case=False)
                          | file_df.columns.str.contains('^.*Europalette.*$', case=False)
                          | file_df.columns.str.contains('^.*Angebotspaket.*$', case=False)
                          | file_df.columns.str.contains('^.*Aufkleber fuer Saeule.*$', case=False)
                          | file_df.columns.str.contains('^.*Aufsetzer.*$', case=False)
                          | file_df.columns.str.contains('^.*Ausstellen der Pflanzen in die Beete  pauschal.*$', case=False)
                          | file_df.columns.str.contains('^.*Ausstellungsflaeche.*$', case=False)
                          | file_df.columns.str.contains('^.*Auswaschen.*$', case=False)
                          | file_df.columns.str.contains('^.*Bild.*$', case=False)
                          | file_df.columns.str.contains('^.*etikette.*$', case=False)
                          )]

Delete specific strings from pandas dataframe with operators chaining

2 Answers2