CODEWITHSUNDEEP
pythonpython-polars
plotmaster473
plotmaster473

Reputation: 160

Select polars columns by index

I have a polars dataframe of species, 89 date columns and 23 unique species. The goal is aggregation by a groupby as well as a range of columns. iloc would be the way to do this in pandas, but the select option doesn't seem to work the way I want it to.

In pandas:

gb = df.groupby(["Common_name"]).agg(dict(zip(df.iloc[:, 32:103].columns, ["mean"] * len(df.iloc[:, 32:103]))))

is there a way to select column indices in polars?

Upvotes: 7

Views: 5873

Answers (1)

Hericks
Hericks

Reputation: 10464

Example data.

import polars as pl

df = pl.DataFrame({
    "a": [1, 2, 3],
    "b": [True, False, False],
    "c": ["a", "b", "c"],
})

Since polars 0.20.26 there exists pl.nth as idiomatic way to select columns by index.

Using a single index.

df.select(pl.nth(0))

shape: (3, 1)
┌─────┐
│ a   │
│ --- │
│ i64 │
╞═════╡
│ 1   │
│ 2   │
│ 3   │
└─────┘

Using a list of indices (introduced in version 0.20.27).

df.select(pl.nth([0, 2]))

shape: (3, 2)
┌─────┬─────┐
│ a   ┆ c   │
│ --- ┆ --- │
│ i64 ┆ str │
╞═════╪═════╡
│ 1   ┆ a   │
│ 2   ┆ b   │
│ 3   ┆ c   │
└─────┴─────┘

Using a range of indices. (introduced in version 0.20.27).

df.select(pl.nth(range(0, 2)))

shape: (3, 2)
┌─────┬───────┐
│ a   ┆ b     │
│ --- ┆ ---   │
│ i64 ┆ bool  │
╞═════╪═══════╡
│ 1   ┆ true  │
│ 2   ┆ false │
│ 3   ┆ false │
└─────┴───────┘

Upvotes: 6

Related Questions