I've seen lots of solutions for coalescing columns together (left to right, etc) But I need to do it row-wise.
| Col A | Col B | Col C | Col D | Col E |
|---|---|---|---|---|
| Row 1 | NA | NA | 1 | text |
| Row 2 | NA | NA | NA | text |
| Row 3 | NA | 4 | 10 | NA |
| Row 4 | 20 | 5 | NA | text |
I want my result to be a single row that looks like...
| Col A | Col B | Col C | Col D | Col E |
|---|---|---|---|---|
| Row 1 | 20 | 4 | 1 | text |
I could iterate over each column, find the first non-NA entry.. then use that as the value. However, I need to do this for hundreds of thousands of distinct partitioned tables for ~30 columns. Surely there's a better solution!
Any help is appreciated :)
Try:
print(df.bfill().head(1))
Prints:
Col A Col B Col C Col D Col E
0 Row 1 20.0 4.0 1.0 text