Simplify scale

[flying-sheep]

scale_sparse has a lot of code that’s not needed.

This PR also fixes zappy compatibility for scale, not that anyone is using it as far as I know.

[codecov]

❌ 1 Tests Failed:

Tests completed	Failed	Passed	Skipped
1675	1	1674	426

View the top 1 failed test(s) by shortest run time

tests/test_preprocessing.py::test_scale

Stack Traces | 0.036s run time

def test_scale():
        adata = pbmc68k_reduced()
        adata.X = adata.raw.X
        v = adata[:, 0 : adata.shape[1] // 2]
        # Should turn view to copy https://github..../anndata/issues/171#issuecomment-508689965
        assert v.is_view
        with pytest.warns(Warning, match="view"):
            sc.pp.scale(v)
        assert not v.is_view
>       assert_allclose(v.X.var(axis=0), np.ones(v.shape[1]), atol=0.01)

#x1B[1m#x1B[31mtests/test_preprocessing.py#x1B[0m:342: 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ 

args = (<function assert_allclose.<locals>.compare at 0x7f5855bce3e0>, matrix([[0.9985714318, 0.9985714246, 0.9985714244, 0.9...1.,
       1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,
       1., 1., 1., 1., 1., 1., 1., 1.]))
kwds = {'equal_nan': True, 'err_msg': '', 'header': 'Not equal to tolerance rtol=1e-07, atol=0.01', 'verbose': True}

    @wraps(func)
    def inner(*args, **kwds):
        with self._recreate_cm():
>           return func(*args, **kwds)
#x1B[1m#x1B[31mE           AssertionError: #x1B[0m
#x1B[1m#x1B[31mE           Not equal to tolerance rtol=1e-07, atol=0.01#x1B[0m
#x1B[1m#x1B[31mE           #x1B[0m
#x1B[1m#x1B[31mE           (shapes (1, 382), (382,) mismatch)#x1B[0m
#x1B[1m#x1B[31mE            x: matrix([[0.998571, 0.998571, 0.998571, 0.998571, 0.998571, 0.998571,#x1B[0m
#x1B[1m#x1B[31mE                    0.998571, 0.998571, 0.998571, 0.998571, 0.998571, 0.998571,#x1B[0m
#x1B[1m#x1B[31mE                    0.998571, 0.998571, 0.998571, 0.998571, 0.998571, 0.998571,...#x1B[0m
#x1B[1m#x1B[31mE            y: array([1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,#x1B[0m
#x1B[1m#x1B[31mE                  1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,#x1B[0m
#x1B[1m#x1B[31mE                  1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1., 1.,...#x1B[0m

#x1B[1m#x1B[.../hostedtoolcache/Python/3.11.11.../x64/lib/python3.11/contextlib.py#x1B[0m:81: AssertionError

To view more test analytics, go to the Test Analytics Dashboard
_{📋 Got 3 mins? Take this short survey to help us improve Test Analytics.}

[scverse-benchmark]

Benchmark changes

Change	Before [329ad66]	After [233d8a3]	Ratio	Benchmark (Parameter)
+	5.48±0.06ms	6.42±0.1ms	1.17	preprocessing_counts.FastSuite.time_calculate_qc_metrics('pbmc68k_reduced', 'counts')
-	6.59±0.4ms	5.08±0.02ms	0.77	preprocessing_counts.FastSuite.time_normalize_total('bmmc', 'counts-off-axis')
+	955M	1.31G	1.37	preprocessing_log.peakmem_scale('pbmc3k', 'off-axis')
+	1.15G	1.5G	1.3	preprocessing_log.peakmem_scale('pbmc3k', None)
+	21.6±0.07ms	27.2±0.2ms	1.26	preprocessing_log.time_highly_variable_genes('pbmc3k', 'off-axis')
-	763±30ms	561±70ms	0.74	preprocessing_log.time_scale('pbmc3k', 'off-axis')

Comparison: https://github.com/scverse/scanpy/compare/329ad666f280b3e7335b3735a04a0a97f430e6f2..233d8a346423bea0043baef72595adf3fddc3d3e
Last changed: Mon, 14 Apr 2025 14:23:48 +0000

More details: https://github.com/scverse/scanpy/pull/3351/checks?check_run_id=40505086063

[flying-sheep]

@Intron7 any idea what I did wrong? Why do the benchmarks say that it uses more memory?

[flying-sheep]

@Intron7 ping

[Intron7]

This would make the default code path for csr even faster. otherwise LGTM