CODEWITHSUNDEEP
pythonpandas
Évariste Galois
Évariste Galois

Reputation: 1033

pandas - copying one column to another without overwriting efficiently

I have columns A and B, where column A rarely has data, and column B is completely full. I want to copy column B to column A and always preserve any value that column A has over column B.

The length of my dataframe is around 1.5 million, so df.iterrows() is a solution, however it would be very time consuming. Is there any optimized pandas function or trick to do this as efficiently as possible?

Upvotes: 1

Views: 777

Answers (2)

Brian
Brian

Reputation: 2286

I would use .loc to get this done

df.loc[df['A'].isnull(), 'A'] = df['B']

This reads like so: located where column 'A' is null and set column 'A' equal to column 'B'.

Upvotes: 2

Denziloe
Denziloe

Reputation: 8132

You can try something like

df["A"][df["A"].isna()] = df["B"][df["A"].isna()]

Edit: the commenter pointed out that there's already a method in pandas for exactly this problem, so use that instead.

Upvotes: 1

Related Questions