I would like a way to get a DataFrame in JSON/dict format excluding NAs efficiently.
This line of code is pretty fast, however when I want to exclude NAs from the results it makes the process very slow.
I have tried different all the different answers here:
The fastest way to achieve this at the moment is this line of code:
[row.dropna().to_dict() for _, row in df.iterrows()]
However these solutions are very slow when dealing with millions of rows
Is there any way to get better time performance?
Source: Python Questions