How to access integer named attribute from pyspark row?-CodePudding

For the row object defined as

from pyspark.sql import Row
rr=Row('a',3)
r = rr(3,4)
print(r)
output:
Row(a=3, 3=4)

I can access the value of a as either r.a or r['a']

But neither can be used to access the value of the key 3. How can I access the value of a Row attribute if it is named by an integer?

CodePudding user response：

You can extract it by using its position within the row. So in your case, that would be the value at index 1:

>>> r[1]
4

Or, if you don't know the position on beforehand, and just want to index using the column name 3 you could do something like this:

# Grabbing index of column with 3 as name
>>> wantedIndex = r.__fields__.index(3)
>>> r[wantedIndex]
4