-1

I'm currently working on a data frame like the one below:

artist week1 week2 week3 week4
Drake 2 2 3 1
Muse NA NA NA NA
Bruno Mars 3 3 4 2
Imagine Dragons NA NA NA NA
Justin Timberlake 2 2 NA 1

What I want to do is to drop the rows that only contain "NA" values. The result should be something like this:

artist week1 week2 week3 week4
Drake 2 2 3 1
Bruno Mars 3 3 4 2
Justin Timberlake 2 2 NA 1

I've tried using the pandas drop() function but drops every row with at least one "NA" value. In that case, the row for Justin Timberlake would be dropped but that's not what I need.

KurtosisCobain
  • 5
  • 2

1 Answers1

0

Use df.dropna() and set how='all' meaning If all values are NA, drop that row or column. then set the subset columns.

df = df.dropna(how='all', subset=['week1', 'week2', 'week3', 'week4'])
print(df)

Or Keep only the rows with at least 2 non-NA values.

df = df.dropna(thresh=2)
print(df)


              artist  week1  week2  week3  week4
0              Drake    2.0    2.0    3.0    1.0
2         Bruno Mars    3.0    3.0    4.0    2.0
4  Justin Timberlake    2.0    2.0    NaN    1.0
Jamiu S.
  • 5,257
  • 5
  • 12
  • 34
  • 1
    Thanks! I have another question though. Is there another way to set the subset when the number of columns is too large to write the name of each one? For example, if I have 300 weeks columns. – KurtosisCobain Feb 03 '23 at 22:40