add_lags timing

time_add_lags.py

import polars as pl
import numpy as np
from sklego.pandas_utils import add_lags
rng = np.random.default_rng(1)
N = 10_000_000
a = rng.integers(0, 10, size=N)
b = rng.integers(0, 10, size=N)
c = rng.integers(0, 10, size=N)
df = pl.DataFrame({'a': a, 'b': b, 'c': c})


In [2]:
   ...: results = %timeit -o add_lags(df, ['a', 'b', 'c'], [1,2,3,4,5])
   ...: results.best
10.8 ms ± 1.11 ms per loop (mean ± std. dev. of 7 runs, 100 loops each)
Out[2]: 0.009629685399995652

In [3]: results = %timeit -o add_lags(df.to_pandas(), ['a', 'b', 'c'], [1,2,3,4,5])
   ...: results.best
   ...:
1.47 s ± 70.4 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
Out[3]: 1.375590296000155