Skip to content

Instantly share code, notes, and snippets.

@tpoterba
Created March 11, 2020 23:36
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save tpoterba/7bded33d0e3f0b4591d07187b245c71d to your computer and use it in GitHub Desktop.
Save tpoterba/7bded33d0e3f0b4591d07187b245c71d to your computer and use it in GitHub Desktop.
$ hail-bench compare 0.2.33-v33-release.json 0.2.33-4a663d2893e7-v34-release-candidate.json
Failed benchmarks in run 1:
table_big_aggregate_compile_and_execute
block_matrix_nested_multiply
pc_relate_big
table_range_array_range_force_count
compile_2k_merge
table_big_aggregate_compilation
Failed benchmarks in run 2:
block_matrix_nested_multiply
pc_relate_big
Benchmark Name Ratio Time 1 Time 2
-------------- ----- ------ ------
split_multi 126.8% 13.270 16.824
shuffle_key_rows_by_mt 117.5% 30.132 35.402
table_group_by_aggregate_unsorted 114.7% 5.260 6.032
union_p10_p100 110.3% 34.296 37.835
write_profile_mt 110.0% 48.703 53.583
ld_prune_profile_25 109.2% 446.966 488.284
table_aggregate_int_stats 108.4% 12.313 13.342
matrix_table_entries_table_no_key 108.2% 45.274 48.999
matrix_table_aggregate_entries 107.5% 5.707 6.135
matrix_table_decode_and_count_just_gt 106.5% 4.286 4.565
matrix_table_call_stats_star_star 104.9% 9.211 9.660
variant_qc 104.6% 8.515 8.909
table_range_means 104.4% 6.473 6.760
table_key_by_shuffle 104.3% 4.889 5.100
table_aggregate_counter 103.7% 11.538 11.967
export_vcf 102.9% 72.092 74.210
matrix_table_many_aggs_row_wise 102.8% 15.133 15.551
python_only_10k_transform 102.4% 91.517 93.701
write_range_table_p10 102.2% 13.004 13.294
join_p10_p100 101.9% 61.948 63.128
join_p10_p1000 101.9% 236.002 240.495
matrix_table_entries_table 101.5% 151.544 153.868
table_aggregate_take_by_strings 101.5% 6.064 6.156
export_range_matrix_table_entry_field_p100 101.4% 9.517 9.651
ndarray_matmul_float64_benchmark 101.3% 3.231 3.273
table_import_strings 101.1% 33.499 33.856
table_group_by_aggregate_sorted 100.6% 5.840 5.877
table_read_force_count_strings 100.4% 4.495 4.514
matrix_multi_write_nothing 100.4% 133.710 134.202
import_vcf_count_rows 100.2% 15.172 15.200
import_bgen_info_score 100.1% 181.273 181.505
export_range_matrix_table_row_p100 100.1% 3.048 3.049
table_range_join_1b_1b 100.0% 1800.000 1800.000
matrix_table_many_aggs_col_wise 99.8% 31.304 31.246
shuffle_order_by_10m_int 99.5% 81.704 81.313
table_annotate_many_nested_dependence 99.5% 6.899 6.863
table_foreign_key_join_same_cardinality 98.9% 18.090 17.884
union_p1000_p10 98.8% 204.554 202.079
join_p100_p10 98.6% 59.777 58.966
table_import_ints 98.6% 121.374 119.650
join_p1000_p1000 98.3% 49.359 48.520
join_p1000_p10 98.1% 233.578 229.227
matrix_table_decode_and_count 98.0% 6.952 6.815
python_only_10k_combine 98.0% 13.206 12.937
import_bgen_filter_count 97.7% 154.769 151.143
write_range_table_p100 97.6% 16.750 16.350
split_multi_hts 97.5% 39.209 38.236
group_by_take_rekey 96.9% 11.600 11.236
table_foreign_key_join_left_higher_cardinality 96.8% 17.060 16.515
table_annotate_many_flat 96.7% 1.008 0.975
matrix_table_nested_annotate_rows_annotate_entries 96.7% 12.790 12.362
union_p1000_p1000 96.4% 18.961 18.272
table_read_force_count_ints 96.2% 8.962 8.619
union_p10_p1000 95.4% 208.367 198.755
variant_and_sample_qc_nested_with_filters_2 95.1% 23.982 22.812
write_range_table_p1000 94.7% 36.835 34.887
kyle_sex_specific_qc 94.7% 9.252 8.762
union_p100_p100 94.7% 16.178 15.313
table_aggregate_downsample_dense 94.6% 67.539 63.919
table_annotate_many_nested_no_dependence 94.6% 3.505 3.314
pc_relate 93.6% 149.803 140.163
table_python_construction 93.3% 1.718 1.602
join_p100_p100 93.1% 46.191 42.990
import_bgen_force_count_all 92.9% 158.120 146.938
hwe_normalized_pca 92.7% 39.690 36.800
matrix_table_filter_entries_unfilter 92.6% 14.295 13.235
write_range_matrix_table_p100 92.5% 6.650 6.148
table_range_force_count 92.1% 5.978 5.509
table_annotate_many_nested_dependence_constants 91.5% 2.782 2.545
variant_and_sample_qc 90.5% 40.853 36.982
linear_regression_rows 89.4% 69.839 62.443
read_force_count_p1000 89.2% 5.657 5.045
ndarray_matmul_int64_benchmark 89.0% 11.205 9.967
matrix_table_scan_count_rows 88.9% 77.026 68.458
sample_qc 88.5% 28.924 25.596
concordance 88.1% 44.367 39.090
import_vcf_write 87.6% 104.785 91.836
table_aggregate_linreg 86.6% 53.932 46.706
read_force_count_p10 86.5% 1.984 1.715
group_by_collect_per_row 85.7% 8.125 6.965
import_bgen_force_count_just_gp 85.4% 171.361 146.358
gnomad_coverage_stats_optimized 85.3% 332.779 283.931
read_with_index_p1000 84.5% 24.392 20.614
read_decode_gnomad_coverage 84.0% 7.036 5.914
genetics_pipeline 84.0% 133.762 112.325
variant_and_sample_qc_nested_with_filters_4 83.6% 46.487 38.878
read_force_count_p100 83.6% 2.438 2.037
matrix_table_filter_entries 83.1% 9.265 7.698
shuffle_key_rows_by_65k_byte_rows 83.1% 35.663 29.627
shuffle_key_rows_by_4096_byte_rows 79.0% 20.717 16.358
matrix_table_array_arithmetic 78.5% 13.818 10.853
per_row_stats_star_star 78.0% 14.358 11.197
large_range_matrix_table_sum 71.2% 589.802 420.065
table_aggregate_downsample_worst_case 68.2% 42.044 28.662
table_aggregate_array_sum 42.1% 21.207 8.933
matrix_table_scan_count_cols 34.3% 2.956 1.014
make_ndarray_bench 5.2% 380.053 19.614
----------------------
Harmonic mean: 78.5%
Geometric mean: 90.9%
Arithmetic mean: 93.8%
Median: 96.8%
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment