CODEWITHSUNDEEP
pythondataframe
charlotte
charlotte

Reputation: 29

Delete dataframe rows based on column values

I have created code based off a website (https://thispointer.com/python-pandas-how-to-drop-rows-in-dataframe-by-conditions-on-column-values/) to delete rows in a data frame based on column values. The column 'zone_type' can have one of 5 values (response_button_text, response_button_image, fixation, timelimit_screen or continue_button). Unless the value for the row is 'response_button_image', I want to remove the row from the data frame.

# select all of the rows that are not to do with the task i.e. fixation screens etc.

indexZoneType = df[ (df['zone_type'] == 'fixation') & (df['zone_type'] == 'response_button_text') & (df['zone_type'] == 'timelimit_screen') & (df['zone_type'] == 'continue_button')].index

# delete these rows

df.drop(indexZoneType , inplace=True)

I thought this code should work? and I get no error, but when executing print(df), the data frame has not changed.

Thanks.

Upvotes: 0

Views: 531

Answers (1)

Georgina Skibinski
Georgina Skibinski

Reputation: 13407

Ok, so your conditions are mutually exclusive, you probably wanted to use 'or' there i.e.

indexZoneType = df[ (df['zone_type'] == 'fixation') |(df['zone_type'] == 'response_button_text') | (df['zone_type'] == 'timelimit_screen') | (df['zone_type'] == 'continue_button')].index

Or even better - using df.isin(...) and df.loc[...]

indexZoneType = df.loc[df['zone_type'].isin(['fixation', 'response_button_text', 'timelimit_screen', 'continue_button'])].index

Or even simpler:

indexZoneType = df.index[df['zone_type'].isin(['fixation', 'response_button_text', 'timelimit_screen', 'continue_button'])]

Reference: https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.isin.html https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.loc.html

Upvotes: 1

Related Questions