Pandas: filter rows with loc on anonymous dataframe

Question

I try to create a query such as
traffic.groupby([round(traffic.a),traffic.r])[["sn","sbr"]].nunique().loc[a==2,:]

The last loc statement does not work (a==2): "name 'a' is not defined.
How can I filter for rows on such an anonymous dataframe?
** Update **
Here is an example
a=[1,2,2,3]
r=[2,2,3,5]
sn=[1,1,2,2]
sbr=[2,2,2,3]

traffic=pd.DataFrame([a,r,sn,sbr],columns=["a","r","sn","sbr"])
traffic.groupby([round(traffic.a),traffic.r])[["sn","sbr"]].nunique().loc[a==2,:]

traffic looks like the following:
    a   r   sn  sbr
0   1   2   2   3
1   2   2   3   5
2   1   1   2   2
3   2   2   2   3

sotmot · Accepted Answer

This might help:
import pandas as pd

a = [1,2,2,3]
r = [2,2,3,5]
sn = [1,1,2,2]
sbr = [2,2,2,3]

traffic = pd.DataFrame([a,r,sn,sbr],columns=["a","r","sn","sbr"])
traffic.groupby([round(traffic.a),traffic.r])[["sn","sbr"]].nunique().T[2].T

The output is:
    sn  sbr
r       
2   2   2

Mokhless · Answer

I wish I could see your dataframe structure! But the way that I  usually query a row without a name (I assume when you're saying "anonymous", your dataset doesn't have headers) is to use iloc.
For example if I want to return the first two rows I would say:
dataframe.iloc[[0, 1]]

Or if I need to return the first two rows for a specific columns let's say column number 4
dataframe.iloc[[0, 1], 3]

I hope this would help!

Pandas: filter rows with loc on anonymous dataframe

2 Answers

Add your own answers!

Ask a Question