Given the following data frame:
import pandas as pd
df = pd.DataFrame({'COL1': ['A', np.nan],
'COL2' : ['A','A']})
df
COL1 COL2
0 A A
1 NaN A
How might I replace the second cell in COL2 with "NaN" (that is, make it null) if the corresponding cell under COL1 is null ("NaN")?
Desired Result:
COL1 COL2
0 A A
1 NaN NaN
Note: I'm looking for a systematic solution that will work across n rows of COL1 and COL2.
Thanks in advance!
You could do this by indexing into the data frame where COL1 is nan:
import pandas as pd
df = pd.DataFrame({'COL1': ['A', np.nan]*100000,
'COL2' : ['A','A']*100000})
df.loc[df.COL1.isnull(), 'COL2'] = np.nan
I used a larger dataframe so that we can compare timings:
%timeit df.loc[df.COL1.isnull(), 'COL2'] = np.nan
100 loops, best of 3: 5.36 ms per loop
Compared to the previous solution which is also a good solution:
%timeit df['COL2'] = np.where(pd.isnull(df['COL1']), np.nan, df['COL2'])
100 loops, best of 3: 10.9 ms per loop
This works:
df['COL2'] = np.where(pd.isnull(df['COL1']), np.nan, df['COL2'])
Is there a preferable way?