I have a dataframe which consists of two columns:

    x   y
0   1   2
1   2   4
2   3   6
3   4   8
4   5  10
5   6  12
6   7  14
7   8  16
8   9  18
9  10  20

I would like to add a column whose value is the index of the first value to meet the following condition: y >= x. For example, for row 2 (x = 3), the first y value greater or equal to 3 is 4 so the output of z for row 2 is (index) 1. I expect the final table to look like:

    x   y   z
0   1   2   0
1   2   4   0
2   3   6   1
3   4   8   1
4   5  10   2
5   6  12   2
6   7  14   3
7   8  16   3
8   9  18   4
9  10  20   4

It should be noted that both x and y are sorted if that should make the solution easier.

I have seen a similar answer but I could not translate it to my situation.

CodePudding user response：

You want np.searchsorted, which assumes df['y'] is sorted:

df['z'] = np.searchsorted(df['y'], df['x'])

Output:

    x   y  z
0   1   2  0
1   2   4  0
2   3   6  1
3   4   8  1
4   5  10  2
5   6  12  2
6   7  14  3
7   8  16  3
8   9  18  4
9  10  20  4