Fix incorrect rank value in data splitting

[yzhangcs]

Fix #6990.

[albertvillanova]

Thanks for the fix.

Could you please add a regression test?

[yzhangcs]

Sure~

[HuggingFaceDocBuilderDev]

The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update.

[lhoestq]

Thanks for the fix !

[github-actions]

Show benchmarks

PyArrow==8.0.0

Show updated benchmarks!

Benchmark: benchmark_array_xd.json

metric	read_batch_formatted_as_numpy after write_array2d	read_batch_formatted_as_numpy after write_flattened_sequence	read_batch_formatted_as_numpy after write_nested_sequence	read_batch_unformated after write_array2d	read_batch_unformated after write_flattened_sequence	read_batch_unformated after write_nested_sequence	read_col_formatted_as_numpy after write_array2d	read_col_formatted_as_numpy after write_flattened_sequence	read_col_formatted_as_numpy after write_nested_sequence	read_col_unformated after write_array2d	read_col_unformated after write_flattened_sequence	read_col_unformated after write_nested_sequence	read_formatted_as_numpy after write_array2d	read_formatted_as_numpy after write_flattened_sequence	read_formatted_as_numpy after write_nested_sequence	read_unformated after write_array2d	read_unformated after write_flattened_sequence	read_unformated after write_nested_sequence	write_array2d	write_flattened_sequence	write_nested_sequence
new / old (diff)	0.005538 / 0.011353 (-0.005815)	0.003997 / 0.011008 (-0.007011)	0.063444 / 0.038508 (0.024935)	0.032552 / 0.023109 (0.009442)	0.266574 / 0.275898 (-0.009324)	0.282841 / 0.323480 (-0.040639)	0.004279 / 0.007986 (-0.003706)	0.002788 / 0.004328 (-0.001540)	0.049226 / 0.004250 (0.044976)	0.044688 / 0.037052 (0.007636)	0.275464 / 0.258489 (0.016975)	0.305278 / 0.293841 (0.011437)	0.030097 / 0.128546 (-0.098450)	0.012237 / 0.075646 (-0.063410)	0.205526 / 0.419271 (-0.213745)	0.036145 / 0.043533 (-0.007388)	0.267395 / 0.255139 (0.012256)	0.289149 / 0.283200 (0.005949)	0.019044 / 0.141683 (-0.122639)	1.162294 / 1.452155 (-0.289861)	1.183642 / 1.492716 (-0.309074)

Benchmark: benchmark_getitem_100B.json

metric	get_batch_of_1024_random_rows	get_batch_of_1024_rows	get_first_row	get_last_row
new / old (diff)	0.139125 / 0.018006 (0.121119)	0.301743 / 0.000490 (0.301253)	0.000260 / 0.000200 (0.000061)	0.000053 / 0.000054 (-0.000001)

Benchmark: benchmark_indices_mapping.json

metric	select	shard	shuffle	sort	train_test_split
new / old (diff)	0.019494 / 0.037411 (-0.017917)	0.063078 / 0.014526 (0.048552)	0.076989 / 0.176557 (-0.099567)	0.121363 / 0.737135 (-0.615773)	0.080040 / 0.296338 (-0.216298)

Benchmark: benchmark_iterating.json

metric	read 5000	read 50000	read_batch 50000 10	read_batch 50000 100	read_batch 50000 1000	read_formatted numpy 5000	read_formatted pandas 5000	read_formatted tensorflow 5000	read_formatted torch 5000	read_formatted_batch numpy 5000 10	read_formatted_batch numpy 5000 1000	shuffled read 5000	shuffled read 50000	shuffled read_batch 50000 10	shuffled read_batch 50000 100	shuffled read_batch 50000 1000	shuffled read_formatted numpy 5000	shuffled read_formatted_batch numpy 5000 10	shuffled read_formatted_batch numpy 5000 1000
new / old (diff)	0.284401 / 0.215209 (0.069192)	2.805397 / 2.077655 (0.727742)	1.555609 / 1.504120 (0.051489)	1.405662 / 1.541195 (-0.135533)	1.459492 / 1.468490 (-0.008999)	0.718376 / 4.584777 (-3.866401)	2.395918 / 3.745712 (-1.349794)	2.976753 / 5.269862 (-2.293108)	1.883938 / 4.565676 (-2.681738)	0.078867 / 0.424275 (-0.345408)	0.005207 / 0.007607 (-0.002400)	0.335178 / 0.226044 (0.109133)	3.313414 / 2.268929 (1.044485)	1.856929 / 55.444624 (-53.587696)	1.565319 / 6.876477 (-5.311158)	1.592723 / 2.142072 (-0.549350)	0.793621 / 4.805227 (-4.011606)	0.134208 / 6.500664 (-6.366456)	0.042853 / 0.075469 (-0.032616)

Benchmark: benchmark_map_filter.json

metric	filter	map fast-tokenizer batched	map identity	map identity batched	map no-op batched	map no-op batched numpy	map no-op batched pandas	map no-op batched pytorch	map no-op batched tensorflow
new / old (diff)	0.981553 / 1.841788 (-0.860235)	11.810438 / 8.074308 (3.736130)	9.529874 / 10.191392 (-0.661518)	0.142216 / 0.680424 (-0.538207)	0.014303 / 0.534201 (-0.519898)	0.304600 / 0.579283 (-0.274684)	0.261869 / 0.434364 (-0.172495)	0.347301 / 0.540337 (-0.193036)	0.437395 / 1.386936 (-0.949541)

PyArrow==latest

Show updated benchmarks!

Benchmark: benchmark_array_xd.json

metric	read_batch_formatted_as_numpy after write_array2d	read_batch_formatted_as_numpy after write_flattened_sequence	read_batch_formatted_as_numpy after write_nested_sequence	read_batch_unformated after write_array2d	read_batch_unformated after write_flattened_sequence	read_batch_unformated after write_nested_sequence	read_col_formatted_as_numpy after write_array2d	read_col_formatted_as_numpy after write_flattened_sequence	read_col_formatted_as_numpy after write_nested_sequence	read_col_unformated after write_array2d	read_col_unformated after write_flattened_sequence	read_col_unformated after write_nested_sequence	read_formatted_as_numpy after write_array2d	read_formatted_as_numpy after write_flattened_sequence	read_formatted_as_numpy after write_nested_sequence	read_unformated after write_array2d	read_unformated after write_flattened_sequence	read_unformated after write_nested_sequence	write_array2d	write_flattened_sequence	write_nested_sequence
new / old (diff)	0.005881 / 0.011353 (-0.005472)	0.004039 / 0.011008 (-0.006969)	0.050241 / 0.038508 (0.011733)	0.032670 / 0.023109 (0.009561)	0.264940 / 0.275898 (-0.010959)	0.287105 / 0.323480 (-0.036374)	0.004844 / 0.007986 (-0.003142)	0.002867 / 0.004328 (-0.001462)	0.048083 / 0.004250 (0.043833)	0.040965 / 0.037052 (0.003913)	0.274390 / 0.258489 (0.015901)	0.312107 / 0.293841 (0.018266)	0.031714 / 0.128546 (-0.096832)	0.012603 / 0.075646 (-0.063043)	0.060698 / 0.419271 (-0.358573)	0.033130 / 0.043533 (-0.010402)	0.264444 / 0.255139 (0.009305)	0.282797 / 0.283200 (-0.000403)	0.027872 / 0.141683 (-0.113811)	1.139026 / 1.452155 (-0.313129)	1.181431 / 1.492716 (-0.311285)

Benchmark: benchmark_getitem_100B.json

metric	get_batch_of_1024_random_rows	get_batch_of_1024_rows	get_first_row	get_last_row
new / old (diff)	0.097314 / 0.018006 (0.079308)	0.301326 / 0.000490 (0.300836)	0.000215 / 0.000200 (0.000015)	0.000049 / 0.000054 (-0.000005)

Benchmark: benchmark_indices_mapping.json

metric	select	shard	shuffle	sort	train_test_split
new / old (diff)	0.023394 / 0.037411 (-0.014018)	0.076270 / 0.014526 (0.061744)	0.089065 / 0.176557 (-0.087491)	0.129996 / 0.737135 (-0.607139)	0.089642 / 0.296338 (-0.206697)

Benchmark: benchmark_iterating.json

metric	read 5000	read 50000	read_batch 50000 10	read_batch 50000 100	read_batch 50000 1000	read_formatted numpy 5000	read_formatted pandas 5000	read_formatted tensorflow 5000	read_formatted torch 5000	read_formatted_batch numpy 5000 10	read_formatted_batch numpy 5000 1000	shuffled read 5000	shuffled read 50000	shuffled read_batch 50000 10	shuffled read_batch 50000 100	shuffled read_batch 50000 1000	shuffled read_formatted numpy 5000	shuffled read_formatted_batch numpy 5000 10	shuffled read_formatted_batch numpy 5000 1000
new / old (diff)	0.295390 / 0.215209 (0.080181)	2.877849 / 2.077655 (0.800194)	1.537129 / 1.504120 (0.033009)	1.409441 / 1.541195 (-0.131754)	1.432468 / 1.468490 (-0.036023)	0.718054 / 4.584777 (-3.866722)	0.930872 / 3.745712 (-2.814841)	2.841028 / 5.269862 (-2.428834)	1.921990 / 4.565676 (-2.643686)	0.077638 / 0.424275 (-0.346637)	0.005494 / 0.007607 (-0.002113)	0.336331 / 0.226044 (0.110287)	3.330490 / 2.268929 (1.061561)	1.887994 / 55.444624 (-53.556630)	1.593332 / 6.876477 (-5.283144)	1.726956 / 2.142072 (-0.415116)	0.783612 / 4.805227 (-4.021615)	0.129926 / 6.500664 (-6.370738)	0.040792 / 0.075469 (-0.034677)

Benchmark: benchmark_map_filter.json

metric	filter	map fast-tokenizer batched	map identity	map identity batched	map no-op batched	map no-op batched numpy	map no-op batched pandas	map no-op batched pytorch	map no-op batched tensorflow
new / old (diff)	0.980274 / 1.841788 (-0.861514)	12.193871 / 8.074308 (4.119563)	10.348934 / 10.191392 (0.157542)	0.141584 / 0.680424 (-0.538840)	0.015737 / 0.534201 (-0.518464)	0.300725 / 0.579283 (-0.278558)	0.127190 / 0.434364 (-0.307174)	0.341142 / 0.540337 (-0.199196)	0.459523 / 1.386936 (-0.927413)

[albertvillanova]

Great! Thanks, @yzhangcs.