Pandas occupies memory when slicing even if no variable is created

[jaymc-betta]

I'm slicing a dataset with loc iloc filter and not saving it to a varaible because I'm exploring a dataset, but it occupies RAM. How can I avoid this? Is it normal? I've never had this issue before, and I work a lot with jupyter in vscode. What's happening?

df.iloc[0:100000] # increases RAM every time I execute it!

I've tried this same code in ipython terminal, and does not increases my RAM usage. So this must be a thing of jupyter+vscode. Any clue?

Edit: I'm using .py files for this (not ipynb) with text selection execution configuration, but I allways do...

[DonJayamanne]

Closing discussion in favour of a duplicate issue #14189