Query column with object datatype

Question

How to do a pandas query when the datatype is object? I tried:

df.query('Train.Test=="Test"')

And also

df.query('Train.Test==Test')

without success.

SampleID,Age,Ctrl.Case,Train.Test,Subject.ID
2241944,22,Ctrl,Train,1
2119485,22,Ctrl,Test,2
2312486,21,Ctrl,Test,5
2813487,19,Ctrl,Test,6
2812488,21,Ctrl,Test,7

Should I convert it to string first? or can I still query with Train.Test as object type?

You will need to do boolean indexing, the `query` method will not work due to the dot(period) in the column name. Try: `df[df['Train.Test'] == 'Test']` — Scott Boston, Oct 20 '22 at 14:45

Code Different · Answer 1 · 2022-10-20T20:14:23.267

3

Any column name that is not a valid Python variable name needs to be quoted with a pair of back ticks:

df.query("`Train.Test` == 'Test'")

edited Oct 20 '22 at 20:14

answered Oct 20 '22 at 14:46

Code Different

90,614
16
144
163

2

Or backtick it! Nice @CodeDifferent I forget about this option. +1 – Scott Boston Oct 20 '22 at 14:47
square-brackets, quotes nothing a backtick. didnt occurred to me :) +1 – Naveed Oct 20 '22 at 14:50

Query column with object datatype

1 Answers1