CODEWITHSUNDEEP
pythonpandas
Amelio Vazquez-Reina
Amelio Vazquez-Reina

Reputation: 96458

Get indices that satisfy some criteria

In Pandas, how can I get a list of indices of a series/dataframe for the entries that satisfy some property?

The following returns a Series object

my_dataframe.loc[:,'some_column'] == 'some_value'

that looks like this:

519     True
509    False
826    False
503    False
511    False
512    False
500    False
507    False
516     True
504    False
521    False
510    False
351    False
522    False
526    False
517    False
501    False

but what I want is two lists, one with

[519, 516]

and one with the rest of indices. How can I do this in Pandas?

Upvotes: 18

Views: 17428

Answers (3)

Hamzah Al-Qadasi
Hamzah Al-Qadasi

Reputation: 9826

Based on @Jeff's example:

df = pd.DataFrame(np.random.rand(10,2),columns=list('AB'))

enter image description here

You can use query instead of a lot of nested brackets:

df.query('B>0.5').index.tolist()    # it prints [0, 7, 8, 9]

Upvotes: 0

TomAugspurger
TomAugspurger

Reputation: 28956

A slight variant on @Jeff's example:

In [18]: df
Out[18]: 
          A         B
0  0.319489  1.012319
1  0.494205 -0.918240
2  1.501922 -0.409661
3 -1.593702  0.705407
4 -0.735312  1.037567
5 -0.201132 -0.673124
6  1.237310 -0.877043
7 -0.946714  0.984164
8 -0.923548  0.415094
9  0.135281 -0.199951

In [14]  list1 =  df.index[df.loc[:, 'B'] > 0]

In [15]: list1
Out[15]: Int64Index([0, 3, 4, 7, 8], dtype=int64)

In [16]: list2 = df.index - list1

In [17]: list2
Out[17]: Int64Index([1, 2, 5, 6, 9], dtype=int64)

Upvotes: 5

Jeff
Jeff

Reputation: 129068

In [8]: df = DataFrame(randn(10,2),columns=list('AB'))

In [9]: df
Out[9]: 
          A         B
0 -1.046978  1.561624
1 -0.264645  0.717171
2  0.112354 -2.084449
3 -1.243482 -1.183749
4  1.055667  0.532444
5 -1.295805  2.168225
6 -1.239725  0.969934
7 -0.354017  1.434943
8 -0.867560  0.810315
9  0.097698 -0.033039

In [10]: df.loc[:,'B'] > 0
Out[10]: 
0     True
1     True
2    False
3    False
4     True
5     True
6     True
7     True
8     True
9    False
Name: B, dtype: bool

In [14]: x = df.loc[:,'B'] > 0

Per Tom/Andy, much simpler 

In [33]: x[x].index
Out[33]: Int64Index([0, 1, 4, 5, 6, 7, 8], dtype=int64)

In [34]: x[~x].index
Out[34]: Int64Index([2, 3, 9], dtype=int64)

Upvotes: 19

Related Questions