CODEWITHSUNDEEP
pythonpandasnumpyseries
nightcrawler
nightcrawler

Reputation: 15

Get list of index of elements of a series in another series

Sorry for the confusing title, but i dont know how to explain this without an example.

I need to get a list of positions of items from one series in another series. The example i was give is the following:

ser1 = pandas.Series([10, 9, 6, 5, 3, 1, 12, 8, 13])
ser2 = pandas.Series([1, 3, 10, 13])

The solution should be [5, 4, 0, 8]

I first thought of using numpy.where(ser1.isin(ser2)) but the array of idex i get is in the wrong order: [0, 4, 5, 8]

If I use ser2.isin(ser1), i get an array of boolean values. Is there a way to get the index of elements instead of this? Or any other way to solve this using other methods?

Upvotes: 1

Views: 598

Answers (1)

jezrael
jezrael

Reputation: 862406

Use broadcasting in numpy for compare with argmax for correct ordering:

out = (ser2.to_numpy() == ser1.to_numpy()[:, None]).argmax(axis=0)
[5, 4, 0, 8]

Solution if some value not matched with np.where and any:

ser1=pd.Series([10, 9, 6, 1, 3, 1, 12, 8, 13])
ser2 = pd.Series([1, 3, 10, 130])

m = ser2.to_numpy() == ser1.to_numpy()[:, None]
out = np.where(m.any(axis=0), m.argmax(axis=0), np.nan)
print (out)
[ 3.  4.  0. nan]

Upvotes: 1

Related Questions