Cannot query dataframe using str.slice with negative indexes

[rwjmiller]

As per this question on SO:

https://stackoverflow.com/questions/74509756/using-pd-query-with-str-slice-with-negative-indexes-and-str-contains

I have the following example dataframe:

df = pd.DataFrame({'a': ['example']})

I can run any of the following without any problems:

df[df['a'].str.slice(0,6).str.contains("e")]  # Works
df.query('a.str.slice(0,6).str.contains("e")')  # Works - same output as above
df[df['a'].str.slice(0,-1).str.contains("e")]  # Works

But not this:

df.query('a.str.slice(0,-1).str.contains("e")')  # Generates an exception!

I get quite a long error message, terminating with

AttributeError: 'UnaryOp' object has no attribute 'value'

Is this perhaps a bug or known issue with Pandas when using the query syntax? I am using Pandas 1.5.1 and python 3.9.7.

[joaoe]

Hi.
Came here to report the same thing

import pandas as pd

pd.DataFrame({"a": ["bc"]}).query("a.str[-1:] == 'c'")
# AttributeError: 'UnaryOp' object has no attribute 'value'

# Correcting to 1
pd.DataFrame({"a": ["bc"]}).query("a.str[1:] == 'c'")
# ValueError: "slice" is not a supported function

Apparently, string slicing operations don't work with query().
This however works df.a.str[-1:] == "c" just fine.