Question about .combine_first()

Question

Let's say I have a data frame df as following:

(Index) Name  Height
0         A     56
1         B     70
2         C       

And a column col as following:

(Index) Height
0
1         69
2         50

What will be the result of df['Height'].combine_first(col) ? There is an overlap at the row with index 1.

Answer 1

It's called "combine_first" because if there are multiple values, it will choose the first one!

Another way of thinking about it is that only null (NaN) values in the first series will be replaced by values from the second series. Non-null values won't be affected. This is basically the way the official docs describe it: https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.combine_first.html