One uses the ‘@’ character in the query string to refer to variables when using pandas.DataFrame.query(). However, doing so seems to throw an error when you refer to class variables. See example below:

In [1]: import seaborn as sns
In [2]: tips = sns.load_dataset('tips')
In [7]: class i:
   ...:     j = 5
   ...: 
In [8]: i.j
Out[8]: 5
In [9]: tips.query("size == @i.j")
Traceback (most recent call last):
  File /usr/local/lib/python3.10/site-packages/pandas/core/computation/scope.py:231 in resolve
    return self.resolvers[key]
  File /usr/local/Cellar/[email protected]/3.10.6_1/Frameworks/Python.framework/Versions/3.10/lib/python3.10/collections/__init__.py:986 in __getitem__
    return self.__missing__(key)            # support subclasses that define __missing__
  File /usr/local/Cellar/[email protected]/3.10.6_1/Frameworks/Python.framework/Versions/3.10/lib/python3.10/collections/__init__.py:978 in __missing__
    raise KeyError(key)
KeyError: 'i'

#Instance variables works correctly as expected


In [19]: m = i()

In [20]: m.j
Out[20]: 5

In [21]: tips.query("size == @m.j")
Out[21]: 
     total_bill   tip     sex smoker   day    time  size
142       41.19  5.00    Male     No  Thur   Lunch     5
155       29.85  5.14  Female     No   Sun  Dinner     5
185       20.69  5.00    Male     No   Sun  Dinner     5
187       30.46  2.00    Male    Yes   Sun  Dinner     5
216       28.15  3.00    Male    Yes   Sat  Dinner     5

Note 1: Class variable used to work. It seems to not now, post upgrade to pandas version 2.2.0

Note 2: I do know that I can use an f-string in the query to circumvent this problem. I’d like to avoid that if I can. It also doesn’t make sense for me to instantiate an object of the class just so that I can use it in the query string.

Has something changed? If so please do point me to any documentation around it.

Thanks!

I’ve tried googling for information around the same and can’t find any.