CODEWITHSUNDEEP
pythonpandas
FooBar
FooBar

Reputation: 16508

Check if set of elements in pandas index

Apparently,

'g' in df.index

is the way to check whether 'g' is in my index. What if I want to do that for many elements? Say, I have an iterable object myIterable (np.ndarray, list), and I want to get the elements that are in or not in my index...

inIterable = [x for x in myIterable if x in df.index]
notInIterable = [x for x in myIterable if x not in df.index]

will work, but is quite inefficient. What is the most efficient way of creating in and notIn?

Upvotes: 2

Views: 1161

Answers (2)

EdChum
EdChum

Reputation: 394189

OK, If I understand correctly to test for membership you can use intersection:

In [132]:

l=[1,2,7]
df = pd.DataFrame({'a':[0,1,2,3,4]})
df.index.intersection(l)
Out[132]:
Int64Index([1, 2], dtype='int64')

For the reverse one method would be to construct a pandas index and use difference:

In [137]:

pd.Index(l).difference(df.index)
Out[137]:
Int64Index([7], dtype='int64')

Upvotes: 3

user2437016
user2437016

Reputation: 120

Use filter:

a = [1,2,3,4,5]    # Index
b = [99,2,99,5]    # MyIterable

in = filter(lambda x: x in a, b)

Note: your example should be:

in = [x for x in myIterable if x in df.index]

And "in" is a reserved word.

Upvotes: 1

Related Questions