Skip to content

perf: more efficient _sparse_nanmean #3570

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
Reovirus wants to merge 27 commits into
base: main
Choose a base branch
from
+74 −17

Rename performance notes file to 3570.perf.md

24f042e
Select commit
Loading
Failed to load commit list.
Open

perf: more efficient _sparse_nanmean #3570

Rename performance notes file to 3570.perf.md
24f042e
Select commit
Loading
Failed to load commit list.
scverse-benchmark / benchmark succeeded Apr 7, 2026 in 1h 37m 14s

Benchmark

Benchmark run successful

Details

All benchmarks:

Change Before [9bc2c1e] After [24f042e] Ratio Benchmark (Parameter)
311M 310M 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('bmmc', 'counts')
312M 311M 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('bmmc', 'counts-off-axis')
4.11G 4.11G 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('lung93k', 'counts')
4.11G 4.11G 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('lung93k', 'counts-off-axis')
378M 378M 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('pbmc3k', 'counts')
377M 378M 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('pbmc3k', 'counts-off-axis')
289M 288M 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('pbmc68k_reduced', 'counts')
289M 289M 1.00 preprocessing_counts.FastSuite.peakmem_calculate_qc_metrics('pbmc68k_reduced', 'counts-off-axis')
314M 314M 1.00 preprocessing_counts.FastSuite.peakmem_log1p('bmmc', 'counts')
314M 314M 1.00 preprocessing_counts.FastSuite.peakmem_log1p('bmmc', 'counts-off-axis')
4.45G 4.45G 1.00 preprocessing_counts.FastSuite.peakmem_log1p('lung93k', 'counts')
4.45G 4.45G 1.00 preprocessing_counts.FastSuite.peakmem_log1p('lung93k', 'counts-off-axis')
385M 385M 1.00 preprocessing_counts.FastSuite.peakmem_log1p('pbmc3k', 'counts')
384M 384M 1.00 preprocessing_counts.FastSuite.peakmem_log1p('pbmc3k', 'counts-off-axis')
288M 288M 1.00 preprocessing_counts.FastSuite.peakmem_log1p('pbmc68k_reduced', 'counts')
288M 288M 1.00 preprocessing_counts.FastSuite.peakmem_log1p('pbmc68k_reduced', 'counts-off-axis')
404M 407M 1.01 preprocessing_counts.FastSuite.peakmem_normalize_total('bmmc', 'counts')
407M 407M 1.00 preprocessing_counts.FastSuite.peakmem_normalize_total('bmmc', 'counts-off-axis')
4.96G 4.96G 1.00 preprocessing_counts.FastSuite.peakmem_normalize_total('lung93k', 'counts')
4.97G 4.97G 1.00 preprocessing_counts.FastSuite.peakmem_normalize_total('lung93k', 'counts-off-axis')
474M 474M 1.00 preprocessing_counts.FastSuite.peakmem_normalize_total('pbmc3k', 'counts')
473M 473M 1.00 preprocessing_counts.FastSuite.peakmem_normalize_total('pbmc3k', 'counts-off-axis')
288M 289M 1.00 preprocessing_counts.FastSuite.peakmem_normalize_total('pbmc68k_reduced', 'counts')
288M 288M 1.00 preprocessing_counts.FastSuite.peakmem_normalize_total('pbmc68k_reduced', 'counts-off-axis')
12.7±0.3ms 12.7±0.3ms 1.00 preprocessing_counts.FastSuite.time_calculate_qc_metrics('bmmc', 'counts')
12.3±0.07ms 12.3±0.06ms 1.00 preprocessing_counts.FastSuite.time_calculate_qc_metrics('bmmc', 'counts-off-axis')
2.06±0s 2.04±0s 0.99 preprocessing_counts.FastSuite.time_calculate_qc_metrics('lung93k', 'counts')
1.65±0.01s 1.62±0s 0.98 preprocessing_counts.FastSuite.time_calculate_qc_metrics('lung93k', 'counts-off-axis')
38.3±0.2ms 37.0±0.7ms 0.96 preprocessing_counts.FastSuite.time_calculate_qc_metrics('pbmc3k', 'counts')
29.2±1ms 27.7±0.9ms 0.95 preprocessing_counts.FastSuite.time_calculate_qc_metrics('pbmc3k', 'counts-off-axis')
4.73±0.07ms 4.76±0.04ms 1.01 preprocessing_counts.FastSuite.time_calculate_qc_metrics('pbmc68k_reduced', 'counts')
4.73±0.08ms 4.79±0.07ms 1.01 preprocessing_counts.FastSuite.time_calculate_qc_metrics('pbmc68k_reduced', 'counts-off-axis')
1.52±0.03ms 1.73±0.1ms ~1.14 preprocessing_counts.FastSuite.time_log1p('bmmc', 'counts')
1.60±0.07ms 1.60±0.09ms 1.00 preprocessing_counts.FastSuite.time_log1p('bmmc', 'counts-off-axis')
634±4ms 638±0.7ms 1.01 preprocessing_counts.FastSuite.time_log1p('lung93k', 'counts')
640±1ms 634±5ms 0.99 preprocessing_counts.FastSuite.time_log1p('lung93k', 'counts-off-axis')
7.10±0.03ms 7.16±0.2ms 1.01 preprocessing_counts.FastSuite.time_log1p('pbmc3k', 'counts')
7.30±0.06ms 7.30±0.02ms 1.00 preprocessing_counts.FastSuite.time_log1p('pbmc3k', 'counts-off-axis')
394±10μs 429±30μs 1.09 preprocessing_counts.FastSuite.time_log1p('pbmc68k_reduced', 'counts')
387±1μs 427±40μs ~1.10 preprocessing_counts.FastSuite.time_log1p('pbmc68k_reduced', 'counts-off-axis')
2.92±8ms 2.60±0.06ms ~0.89 preprocessing_counts.FastSuite.time_normalize_total('bmmc', 'counts')
6.74±0.2ms 6.66±0.2ms 0.99 preprocessing_counts.FastSuite.time_normalize_total('bmmc', 'counts-off-axis')
545±5ms 538±9ms 0.99 preprocessing_counts.FastSuite.time_normalize_total('lung93k', 'counts')
2.85±0.01s 2.78±0.02s 0.98 preprocessing_counts.FastSuite.time_normalize_total('lung93k', 'counts-off-axis')
8.56±0.2ms 8.61±0.4ms 1.01 preprocessing_counts.FastSuite.time_normalize_total('pbmc3k', 'counts')
35.6±6ms 34.0±0.2ms 0.96 preprocessing_counts.FastSuite.time_normalize_total('pbmc3k', 'counts-off-axis')
549±1μs 547±0.9μs 1.00 preprocessing_counts.FastSuite.time_normalize_total('pbmc68k_reduced', 'counts')
547±1μs 547±3μs 1.00 preprocessing_counts.FastSuite.time_normalize_total('pbmc68k_reduced', 'counts-off-axis')
399M 400M 1.00 preprocessing_counts.PreprocessingCountsRngSuite.peakmem_downsample_per_cell('pbmc3k', 'random_state')
399M 399M 1.00 preprocessing_counts.PreprocessingCountsRngSuite.peakmem_downsample_per_cell('pbmc3k', 'rng')
336M 337M 1.00 preprocessing_counts.PreprocessingCountsRngSuite.peakmem_downsample_per_cell('pbmc68k_reduced', 'random_state')
336M 336M 1.00 preprocessing_counts.PreprocessingCountsRngSuite.peakmem_downsample_per_cell('pbmc68k_reduced', 'rng')
468M 468M 1.00 preprocessing_counts.PreprocessingCountsRngSuite.peakmem_downsample_total('pbmc3k', 'random_state')
432M 432M 1.00 preprocessing_counts.PreprocessingCountsRngSuite.peakmem_downsample_total('pbmc3k', 'rng')
341M 341M 1.00 preprocessing_counts.PreprocessingCountsRngSuite.peakmem_downsample_total('pbmc68k_reduced', 'random_state')
338M 338M 1.00 preprocessing_counts.PreprocessingCountsRngSuite.peakmem_downsample_total('pbmc68k_reduced', 'rng')
205±1ms 207±3ms 1.01 preprocessing_counts.PreprocessingCountsRngSuite.time_downsample_per_cell('pbmc3k', 'random_state')
72.7±0.8ms 73.3±0.5ms 1.01 preprocessing_counts.PreprocessingCountsRngSuite.time_downsample_per_cell('pbmc3k', 'rng')
25.4±0.09ms 25.1±0.3ms 0.99 preprocessing_counts.PreprocessingCountsRngSuite.time_downsample_per_cell('pbmc68k_reduced', 'random_state')
16.2±0.06ms 16.7±0.06ms 1.03 preprocessing_counts.PreprocessingCountsRngSuite.time_downsample_per_cell('pbmc68k_reduced', 'rng')
194±10ms 190±1ms 0.98 preprocessing_counts.PreprocessingCountsRngSuite.time_downsample_total('pbmc3k', 'random_state')
63.1±2ms 65.9±6ms 1.04 preprocessing_counts.PreprocessingCountsRngSuite.time_downsample_total('pbmc3k', 'rng')
12.8±0.5ms 13.2±0.08ms 1.03 preprocessing_counts.PreprocessingCountsRngSuite.time_downsample_total('pbmc68k_reduced', 'random_state')
7.06±0.4ms 7.12±0.3ms 1.01 preprocessing_counts.PreprocessingCountsRngSuite.time_downsample_total('pbmc68k_reduced', 'rng')
431M 430M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_cells('pbmc3k', 'counts')
431M 431M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_cells('pbmc3k', 'counts-off-axis')
299M 300M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_cells('pbmc68k_reduced', 'counts')
299M 299M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_cells('pbmc68k_reduced', 'counts-off-axis')
431M 430M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_genes('pbmc3k', 'counts')
431M 431M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_genes('pbmc3k', 'counts-off-axis')
299M 299M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_genes('pbmc68k_reduced', 'counts')
299M 299M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_filter_genes('pbmc68k_reduced', 'counts-off-axis')
1.11G 1.1G 0.99 preprocessing_counts.PreprocessingCountsSuite.peakmem_scrublet('pbmc3k', 'counts')
1.11G 1.1G 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_scrublet('pbmc3k', 'counts-off-axis')
530M 528M 0.99 preprocessing_counts.PreprocessingCountsSuite.peakmem_scrublet('pbmc68k_reduced', 'counts')
529M 528M 1.00 preprocessing_counts.PreprocessingCountsSuite.peakmem_scrublet('pbmc68k_reduced', 'counts-off-axis')
61.1±9ms 56.2±0.8ms 0.92 preprocessing_counts.PreprocessingCountsSuite.time_filter_cells('pbmc3k', 'counts')
59.1±0.4ms 58.6±0.5ms 0.99 preprocessing_counts.PreprocessingCountsSuite.time_filter_cells('pbmc3k', 'counts-off-axis')
10.5±1ms 10.1±0.7ms 0.96 preprocessing_counts.PreprocessingCountsSuite.time_filter_cells('pbmc68k_reduced', 'counts')
10.0±0.7ms 10.3±1ms 1.03 preprocessing_counts.PreprocessingCountsSuite.time_filter_cells('pbmc68k_reduced', 'counts-off-axis')
52.7±0.5ms 51.5±0.3ms 0.98 preprocessing_counts.PreprocessingCountsSuite.time_filter_genes('pbmc3k', 'counts')
51.9±0.3ms 51.0±0.8ms 0.98 preprocessing_counts.PreprocessingCountsSuite.time_filter_genes('pbmc3k', 'counts-off-axis')
11.6±2ms 10.1±0.8ms ~0.87 preprocessing_counts.PreprocessingCountsSuite.time_filter_genes('pbmc68k_reduced', 'counts')
10.2±0.8ms 10.4±1ms 1.02 preprocessing_counts.PreprocessingCountsSuite.time_filter_genes('pbmc68k_reduced', 'counts-off-axis')
2.76±0.01s 5.66±2s ~2.05 preprocessing_counts.PreprocessingCountsSuite.time_scrublet('pbmc3k', 'counts')
2.64±0.03s 2.56±0.03s 0.97 preprocessing_counts.PreprocessingCountsSuite.time_scrublet('pbmc3k', 'counts-off-axis')
554±4ms 556±5ms 1.00 preprocessing_counts.PreprocessingCountsSuite.time_scrublet('pbmc68k_reduced', 'counts')
546±10ms 562±5ms 1.03 preprocessing_counts.PreprocessingCountsSuite.time_scrublet('pbmc68k_reduced', 'counts-off-axis')
441M 441M 1.00 preprocessing_log.PreprocessingSuite.peakmem_highly_variable_genes('pbmc3k', 'off-axis')
492M 491M 1.00 preprocessing_log.PreprocessingSuite.peakmem_highly_variable_genes('pbmc3k', None)
295M 295M 1.00 preprocessing_log.PreprocessingSuite.peakmem_highly_variable_genes('pbmc68k_reduced', 'off-axis')
297M 298M 1.00 preprocessing_log.PreprocessingSuite.peakmem_highly_variable_genes('pbmc68k_reduced', None)
569M 573M 1.01 preprocessing_log.PreprocessingSuite.peakmem_pca('pbmc3k', 'off-axis')
587M 588M 1.00 preprocessing_log.PreprocessingSuite.peakmem_pca('pbmc3k', None)
496M 492M 0.99 preprocessing_log.PreprocessingSuite.peakmem_pca('pbmc68k_reduced', 'off-axis')
498M 499M 1.00 preprocessing_log.PreprocessingSuite.peakmem_pca('pbmc68k_reduced', None)
n/a n/a n/a preprocessing_log.PreprocessingSuite.peakmem_regress_out('pbmc3k', 'off-axis')
n/a n/a n/a preprocessing_log.PreprocessingSuite.peakmem_regress_out('pbmc3k', None)
350M 350M 1.00 preprocessing_log.PreprocessingSuite.peakmem_regress_out('pbmc68k_reduced', 'off-axis')
353M 353M 1.00 preprocessing_log.PreprocessingSuite.peakmem_regress_out('pbmc68k_reduced', None)
1.3G 1.3G 1.00 preprocessing_log.PreprocessingSuite.peakmem_scale('pbmc3k', 'off-axis')
1.5G 1.5G 1.00 preprocessing_log.PreprocessingSuite.peakmem_scale('pbmc3k', None)
337M 336M 1.00 preprocessing_log.PreprocessingSuite.peakmem_scale('pbmc68k_reduced', 'off-axis')
339M 339M 1.00 preprocessing_log.PreprocessingSuite.peakmem_scale('pbmc68k_reduced', None)
34.6±0.5ms 34.5±2ms 1.00 preprocessing_log.PreprocessingSuite.time_highly_variable_genes('pbmc3k', 'off-axis')
40.9±1ms 40.1±0.6ms 0.98 preprocessing_log.PreprocessingSuite.time_highly_variable_genes('pbmc3k', None)
16.6±0.3ms 16.7±0.08ms 1.00 preprocessing_log.PreprocessingSuite.time_highly_variable_genes('pbmc68k_reduced', 'off-axis')
16.7±0.08ms 16.6±0.3ms 1.00 preprocessing_log.PreprocessingSuite.time_highly_variable_genes('pbmc68k_reduced', None)
1.96±0.01s 1.93±0.01s 0.98 preprocessing_log.PreprocessingSuite.time_pca('pbmc3k', 'off-axis')
2.15±0.02s 2.14±0.02s 0.99 preprocessing_log.PreprocessingSuite.time_pca('pbmc3k', None)
165±8ms 163±20ms 0.99 preprocessing_log.PreprocessingSuite.time_pca('pbmc68k_reduced', 'off-axis')
167±6ms 144±40ms ~0.86 preprocessing_log.PreprocessingSuite.time_pca('pbmc68k_reduced', None)
n/a n/a n/a preprocessing_log.PreprocessingSuite.time_regress_out('pbmc3k', 'off-axis')
n/a n/a n/a preprocessing_log.PreprocessingSuite.time_regress_out('pbmc3k', None)
17.1±0.5ms 16.7±0.4ms 0.97 preprocessing_log.PreprocessingSuite.time_regress_out('pbmc68k_reduced', 'off-axis')
16.8±0.6ms 16.8±0.8ms 1.00 preprocessing_log.PreprocessingSuite.time_regress_out('pbmc68k_reduced', None)
503±1ms 503±4ms 1.00 preprocessing_log.PreprocessingSuite.time_scale('pbmc3k', 'off-axis')
546±1ms 546±0.7ms 1.00 preprocessing_log.PreprocessingSuite.time_scale('pbmc3k', None)
4.71±0.1ms 4.71±0.1ms 1.00 preprocessing_log.PreprocessingSuite.time_scale('pbmc68k_reduced', 'off-axis')
4.79±0.07ms 4.79±0.08ms 1.00 preprocessing_log.PreprocessingSuite.time_scale('pbmc68k_reduced', None)
288M 287M 1.00 tools.ToolsSuite.peakmem_diffmap
294M 294M 1.00 tools.ToolsSuite.peakmem_leiden
375M 374M 1.00 tools.ToolsSuite.peakmem_rank_genes_groups
+ 290M 338M 1.16 tools.ToolsSuite.peakmem_score_genes
450M 449M 1.00 tools.ToolsSuite.peakmem_umap
16.7±0.7ms 16.3±0.4ms 0.98 tools.ToolsSuite.time_diffmap
19.7±0.09ms 19.8±0.1ms 1.00 tools.ToolsSuite.time_leiden
60.1±6ms 60.3±5ms 1.00 tools.ToolsSuite.time_rank_genes_groups
- 19.4±0.2ms 17.5±0.1ms 0.90 tools.ToolsSuite.time_score_genes
1.02±0s 1.02±0s 1.00 tools.ToolsSuite.time_umap
View more details on scverse-benchmark