How to drop row in pandas DataFrame by row index number-CodePudding

I want to drop some rows with does match with if conditions in my code, but it can not , it pop out the error as below , could you please help assist ?

def listconvert(df_test):
    df1 = []
    for i in df_test:
        if str(i) != 'nan':
            df1.append(i)
    return df1


for j in range(len(df)-1):
    # print(len(listconvert(df.iloc[j].values)) )
    if len(listconvert(df.iloc[j].values)) < len(df.columns):
        # df.drop(df.loc[df.index==j].index,inplace=True)
        print(len(listconvert(df.iloc[j].values)))
        df = df.drop(df.index[j])
    else:
        pass
# df1.to_excel("Mar-131.xlsx")

the error as below:

---------------------------------------------------------------------------
IndexError                                Traceback (most recent call last)
~\AppData\Local\Temp/ipykernel_16140/2461102218.py in <module>
     45 for j in range(len(df)-1):
     46     # print(len(listconvert(df.iloc[j].values)) )
---> 47     if len(listconvert(df.iloc[j].values)) < len(df.columns):
     48         # df.drop(df.loc[df.index==j].index,inplace=True)
     49         print(len(listconvert(df.iloc[j].values)))

IndexError: single positional indexer is out-of-bounds

CodePudding user response：

You should rethink your code. Fundamentally all it's doing is dropping rows which contain one or more NANs. You can do that 1000x faster like this:

df.dropna(inplace=True)

Or, just for inspiration to show how you could do it yourself if there weren't a prebuilt function for it:

df = df[df.notna().all(axis=1)]

That's it, that replaces all the code you had, with no slow for loops.