CODEWITHSUNDEEP
pythonpandasdataframenan
dinn_
dinn_

Reputation: 93

Is there a way to find rows with NaN and then drop that as well as n number of rows after?

I have a spreadsheet where data is laid out in a strange way with chunks of data separated by a single empty row of NaN cells. I can remove the NaN rows with drop.na of course, however is there a way for me to drop those NaN rows as well as a specified number of rows beneath them?

For example I'd like to drop each of the NaN rows in the dataframe as well as the 2 rows below it that are not NaN.

Upvotes: 1

Views: 98

Answers (1)

SultanOrazbayev
SultanOrazbayev

Reputation: 16561

One way is to create a mask from shifted column values, for example:

# identify nan rows, true if nan
df['row_is_na'] = df['some_column'].isna()

# identify two rows that follow nan, true if after nan
df['rows_after_nan'] = df['row_is_na'].shift(-1) | df['row_is_na'].shift(-2)

# apply the mask
df = df[~df['rows_after_nan']]

Upvotes: 1

Related Questions