Describe the bug

A validation check for the precomputed gram matrix has been introduced in version 1.0.0 (#19004).

This check sometimes misleadingly fails when the matrix has dtype float32 and the arbitrary selected feature columns are sparse.

Code snippet to reproduce attached.

I could add a pr in the following days to fix that if wanted.

Steps/Code to Reproduce

from sklearn.linear_model import LassoCV
import numpy as np

m = LassoCV()

np.random.seed(seed=3)

X = np.random.random((10000, 50)).astype(np.float32)
X[:, 25] = np.where(X[:, 25] < 0.98, 0, 1)
X[:, 26] = np.where(X[:, 26] < 0.98, 0, 1)
y = np.random.random((10000, 1)).astype(np.float32)

m.fit(X, y)

Expected Results

No Exception thrown

Actual Results

ValueError: Gram matrix passed in via 'precompute' parameter did not pass validation when a single element was checked - please check that it was computed properly. For element (25,26) we computed -0.4163646101951599 but the user-supplied value was -0.41635191440582275.

Versions

1.0.1