What is the python equivalent of this in operator? I am trying to filter down a pandas database by having rows only remain if a column in the row has a value found in my list.
I tried using any() and am having immense difficulty with this.
- 1@Jeff post that as an answer.James Mertz– James Mertz2014-08-08 15:00:03 +00:00Commented Aug 8, 2014 at 15:00
- That's beautiful, exactly what I was looking for. You know how hard it is to google "in" and special symbols.wolfsatthedoor– wolfsatthedoor2014-08-08 15:01:21 +00:00Commented Aug 8, 2014 at 15:01
- I don't see the difficulty. Googling "pandas in operator" provides pandas.pydata.org/pandas-docs/stable/indexing.html as first hit and a text search of "in operator" on that page let's you immediately find what you are looking for.K.-Michael Aye– K.-Michael Aye2014-08-08 21:21:47 +00:00Commented Aug 8, 2014 at 21:21
- I googled Python rather than pandas, I didn't know it was a Pandas specific thing.wolfsatthedoor– wolfsatthedoor2014-08-08 21:28:51 +00:00Commented Aug 8, 2014 at 21:28
Add a comment |
4 Answers
Pandas comparison with R docs are here.
s <- 0:4 s %in% c(2,4) The isin method is similar to R %in% operator:
In [13]: s = pd.Series(np.arange(5),dtype=np.float32) In [14]: s.isin([2, 4]) Out[14]: 0 False 1 False 2 True 3 False 4 True dtype: bool 
Comments
FWIW: without having to call pandas, here's the answer using a for loop and list compression in pure python
x = [2, 3, 5] y = [1, 2, 3] # for loop for i in x: [].append(i in y) Out: [True, True, False] # list comprehension [i in y for i in x] Out: [True, True, False] 
Comments
If you want to use only numpy without panads (like a use case I had) then you can:
import numpy as np x = np.array([1, 2, 3, 10]) y = np.array([10, 11, 2]) np.isin(y, x) This is equivalent to:
c(10, 11, 2) %in% c(1, 2, 3, 10) Note that the last line will work only for numpy >= 1.13.0, for older versions you'll need to use np.in1d.
Comments
As others indicate, in operator of base Python works well.
myList = ["a00", "b000", "c0"] "a00" in myList # True "a" in myList # False 
1 Comment
russellpierce
But requires an atomic left side to yield results that match R's
%in% in calling semantics. E.g. ["a00", "node", "c0"] in myList isn't what someone used to %in% is going to expect.