What's the polars equivalent to the pandas `.iloc` method?

Question

I'm looking for the recommended way to select an individual row of a polars.DataFrame by row number: something largely equivalent to pandas.DataFrame's .iloc[[n]] method for a given integer n.

For polars imported as pl and a polars DataFrame df, my current approach would be:

# for example
n = 3

# create row index, filter for individual row, drop the row index.
new_df = (
    df.with_row_index()
    .filter(pl.col('index') == n)
    .select(pl.exclude('index'))
)

I'm migrating from Pandas, and I've read the Pandas-to-Polars migration guide, but a slick solution to this specific case wasn't addressed there. Edit: to clarify, I am looking for an approach that returns a polars.DataFrame object for the chosen row.

Does anyone have something slicker?

It would probably depend on what you're doing with the row. For "quick", "interactive" usage df[3] works - but wouldn't be "recommended" if you were composing larger expressions. — jqurious
– jqurious, Commented Jan 25, 2024 at 17:55
It seems the Polars does not really use an index. The focus in on the data and not the index. — Cam
– Cam, Commented Jan 25, 2024 at 18:06
But you can set a column as the index via polars.DataFrame.with_row_index — Cam
– Cam, Commented Jan 25, 2024 at 18:14

Talha Tayyab · Accepted Answer · 2024-01-25 17:55:20Z

26

This is a very good sheet by: @Liam Brannigan

Credit to them.

https://www.rhosignal.com/posts/polars-pandas-cheatsheet/

A glimse from the sheet:

You can find other information related to Filtering Rows using iloc and its equivalent in polars in the sheet.

answered Jan 25, 2024 at 17:55

Talha Tayyab

30.7k55 gold badges139 silver badges170 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Ro.oT · Accepted Answer · 2024-01-25 17:55:31Z

1

I guess, the direct equivalent of .iloc from pandas is the .row() method in polars.

If you have a dataframe, df:

df = pl.DataFrame(
    {
        "foo": [1, 2, 3],
        "bar": [6, 7, 8],
        "ham": ["a", "b", "c"],
    }
)

Here's how you'd access a row:

df.row(2)
(3, 8, 'c')

answered Jan 25, 2024 at 17:55

Ro.oT

2,1147 gold badges19 silver badges29 bronze badges

2 Comments

montol Over a year ago

Thanks for sharing this - I was looking for an approach that returns a dataframe object - I'll clarify in the question.

Ro.oT Over a year ago

Oh I'm sorry. I didn't understand that from the original question.

Collectives™ on Stack Overflow

What's the polars equivalent to the pandas `.iloc` method?

2 Answers 2

Comments

2 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

2 Comments

Your Answer

Sign up or log in

Post as a guest

Related