PySpark: dynamic union of DataFrames with different columns

Question

Consider the arrays as shown here. I have 3 sets of array:

Array 1:

C1  C2  C3
1   2   3
9   5   6

Array 2:

C2 C3 C         


        

Answer 1

I would try

df = df1.join(df2, ['each', 'shared', 'col], how='full')