Allow use on 1d arrays

[flying-sheep]

One thing we could do is making the assert_array_equals strict=True.

That’d require downcasting float64 to float32 in some cases though.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 98.95%. Comparing base (a1643f9) to head (6bcd3b7).
Report is 1 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #84      +/-   ##
==========================================
+ Coverage   98.91%   98.95%   +0.03%     
==========================================
  Files          17       17              
  Lines         370      381      +11     
==========================================
+ Hits          366      377      +11     
  Misses          4        4              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[codspeed-hq]

CodSpeed Performance Report

Merging #84 will improve performances by ×4

_{Comparing 0d (6bcd3b7) with main (a1643f9)}

Summary

⚡ 18 improvements
✅ 106 untouched benchmarks
🆕 36 new benchmarks

Benchmarks breakdown

	Benchmark	BASE	HEAD	Change
🆕	test_stats_benchmark[1d-all-numpy.ndarray-float32-is_constant]	N/A	2.7 ms	N/A
🆕	test_stats_benchmark[1d-all-numpy.ndarray-float32-mean]	N/A	2.9 ms	N/A
🆕	test_stats_benchmark[1d-all-numpy.ndarray-float32-mean_var]	N/A	8.9 ms	N/A
🆕	test_stats_benchmark[1d-all-numpy.ndarray-float32-sum]	N/A	2.8 ms	N/A
🆕	test_stats_benchmark[1d-all-numpy.ndarray-float64-is_constant]	N/A	4.8 ms	N/A
🆕	test_stats_benchmark[1d-all-numpy.ndarray-float64-mean]	N/A	4.7 ms	N/A
🆕	test_stats_benchmark[1d-all-numpy.ndarray-float64-mean_var]	N/A	17.1 ms	N/A
🆕	test_stats_benchmark[1d-all-numpy.ndarray-float64-sum]	N/A	4.7 ms	N/A
🆕	test_stats_benchmark[1d-all-numpy.ndarray-int32-is_constant]	N/A	2.8 ms	N/A
🆕	test_stats_benchmark[1d-all-numpy.ndarray-int32-mean]	N/A	4.1 ms	N/A
🆕	test_stats_benchmark[1d-all-numpy.ndarray-int32-mean_var]	N/A	8.9 ms	N/A
🆕	test_stats_benchmark[1d-all-numpy.ndarray-int32-sum]	N/A	4.1 ms	N/A
🆕	test_stats_benchmark[1d-all-scipy.sparse.csc_array-float32-is_constant]	N/A	2.7 ms	N/A
🆕	test_stats_benchmark[1d-all-scipy.sparse.csc_array-float32-mean]	N/A	2.9 ms	N/A
🆕	test_stats_benchmark[1d-all-scipy.sparse.csc_array-float32-mean_var]	N/A	14.7 ms	N/A
🆕	test_stats_benchmark[1d-all-scipy.sparse.csc_array-float32-sum]	N/A	2.8 ms	N/A
🆕	test_stats_benchmark[1d-all-scipy.sparse.csc_array-float64-is_constant]	N/A	4.8 ms	N/A
🆕	test_stats_benchmark[1d-all-scipy.sparse.csc_array-float64-mean]	N/A	4.7 ms	N/A
🆕	test_stats_benchmark[1d-all-scipy.sparse.csc_array-float64-mean_var]	N/A	21.4 ms	N/A
🆕	test_stats_benchmark[1d-all-scipy.sparse.csc_array-float64-sum]	N/A	4.6 ms	N/A
...	...	...	...	...

ℹ️ Only the first 20 benchmarks are displayed. Go to the app to view all benchmarks.

[flying-sheep]

without this, 1D csr_arrays sum up to np.float64 no matter if bool, int, or float32.

that’s not the case for 2D ones, so I should probably report this as an issue to scipy.