2c:null 2d:["$","title",null,{"children":"pypocketmap benchmarks | Dylan Burati's Workspace | Observable"}] 32:T66a,NB: the maximum load factor is 2/3 for `dict` ([link](https://github.com/python/cpython/blob/3.10/Objects/dictobject.c#L413)) and 79/100 for a `pypocketmap`. At 37 seconds, the spike in `dict` memory usage corresponds to the backing arrays growing to 2²⁶ (≈ 67 million) slots. The `pypocketmap` stays at half that number of slots, because the final size of the map was ${sizes[2].toLocaleString()} = ${Math.round(1000 * sizes[2] / (2**25)) / 1000}(2²⁵). This happens for about a quarter of all sizes, although the fraction doesn't converge nicely.

Open for weird math

for all k, the range \[(1/3) 2^k, (2/3) 2^k] has `dict_slots = 2^k`
- the portion of the range before (0.79/2) 2^k has `pypocketmap_slots = 2^(k-1)`, and the rest has `2^k`
- (0.395 - 1/3) * 3 = ${Math.round(10000 * (0.395 - 1/3) * 3) / 100}%
for all k, the range \[(0.79/2) 2^k, 0.79 2^k] has `pypocketmap_slots = 2^k`
- the portion of the range after (2/3) 2^k has `dict_slots = 2^(k+1)` and the rest has `2^k`
- (0.79 - 2/3) / 0.395 = ${Math.round(10000 * (0.79 - 2/3) / 0.395) / 100}%
for all N, lg `dict_slots` = ceil(lg 1.5N) = ceil(lg N + lg 1.5)
- lg `pypocketmap_slots` = ceil(lg 100N/79) = ceil(lg N + lg 100/79)
- ceil(lg N + lg 1.5) > ceil(lg N + lg 100/79) when the fractional part of lg N is in the range \[1 − lg 1.5, 1 − lg 100/79]
- 2/(100/79) - 2/(3/2) = ${Math.round(10000 * 2 * (0.79 - 2/3)) / 100}%

33:Tb73,allProcessed = { const runGroups = benchmarks[radios]?.proc; if (runGroups == null) { return []; } const rename = {mdict: "pypocketmap"} const runLogs = runGroups .flatMap(({ implementation, ...rest }) => Object.entries(rest).map(([size, data]) => ({ implementation: rename[implementation] || implementation, size, data, }) )); const iterLogs = runLogs.flatMap(({ data, ...rest }) => { const dataSplit = data.reduce((acc, cur) => { acc[acc.length - 1].push(cur); if (cur.elapsed !== undefined) { acc.push([]); } return acc; }, [[]]); return dataSplit.filter((d) => d.length > 0).map((data) => ({ data, ...rest })); }); const iterLogsDiff = iterLogs.map(rl => { const [d, _, __, maxMem, threads] = rl.data.slice(0, -1).reduce(([acc, prevTime, prevCPU, cmaxMem, threads], cur) => { const { time_seconds, start_time_seconds, cpu_seconds_total, thread_cpu_seconds_total, resident_memory_bytes } = cur; const threadCPU = []; for (let [tid, amt] of thread_cpu_seconds_total) { threadCPU.push([tid, amt - (threads.get(tid) || 0)]); threads.set(tid, amt); } const elt = { implementation: rl.implementation, size: rl.size, timestamp: time_seconds - start_time_seconds, timeDiff: time_seconds - prevTime, cpu: cpu_seconds_total - prevCPU, mem: resident_memory_bytes / 1024 / 1024 / 1024, threadCPU, } return [[...acc, elt], time_seconds, cpu_seconds_total, Math.max(cmaxMem, elt.mem), threads]; }, [[], 0, 0, 0, new Map()]); const mainThread = Array.from(threads).sort(([_1, v1], [_2, v2]) => v2 - v1)[0][0]; return [ d.map(({ threadCPU, ...elt }) => ({ ...elt, mainCPU: threadCPU.filter(([k]) => k === mainThread).reduce((acc, [_, v]) => acc + v, 0) / elt.timeDiff * 100, backCPU: threadCPU.filter(([k]) => k !== mainThread).reduce((acc, [_, v]) => acc + v, 0) / elt.timeDiff * 100, })), { implementation: rl.implementation, size: rl.size, time: rl.data[rl.data.length - 1].elapsed, stdout: rl.data[rl.data.length - 1].stdout.data, memory: maxMem, }, ]; }); const summaries = iterLogsDiff.map(([_, summary]) => summary); // use the median-length run as the graph example const timeByImplSize = summaries.map((e) => [`${e.implementation}-${e.size}`, e.time]) .reduce((acc, [k, v], i) => { acc.set(k, [...(acc.get(k) || []), [i, v]]); return acc; }, new Map()); for (const [_, lst] of timeByImplSize) { lst.sort((a, b) => a[1] - b[1]); } return [ iterLogsDiff .filter(([_, e], i) => i === timeByImplSize.get(`${e.implementation}-${e.size}`)[1][0]) .flatMap(([recording, _]) => recording), summaries, ]; }34:[{"id":"c3186bf1052b6d435c1fa7fe0af55777f27c4740ea2880ea3038ac07e1afd897e20783335b95b5da2da243411fc15329e7bb598e0898a9c8bbeb4df143c7bcd9","href":"https://static.observableusercontent.com/files/c3186bf1052b6d435c1fa7fe0af55777f27c4740ea2880ea3038ac07e1afd897e20783335b95b5da2da243411fc15329e7bb598e0898a9c8bbeb4df143c7bcd9","name":"proc_py310-linux_dict_md@3.ldjson","type":"application/octet-stream","lastModified":1708211967066.319,"size":24691},{"id":"3bee4c9d53211f7856ce179232a9c3da8ed1ba8f3ad52fcfd84462e368309654c14130e907e7645e455e191e8dd89483844bcbf22dd15878e27ac0e94c742faa","href":"https://static.observableusercontent.com/files/3bee4c9d53211f7856ce179232a9c3da8ed1ba8f3ad52fcfd84462e368309654c14130e907e7645e455e191e8dd89483844bcbf22dd15878e27ac0e94c742faa","name":"proc_py310-linux_mdict_md@3.ldjson","type":"application/octet-stream","lastModified":1708211967084.868,"size":27401},{"id":"13017ed24987d40915c40d096a966a6c7478a1a2ee14f015bc56960a0f68c2bb63b31da15e53e96a4a9aa8b78fc5051e4d308ca9032210147faf3836b3f180a6","href":"https://static.observableusercontent.com/files/13017ed24987d40915c40d096a966a6c7478a1a2ee14f015bc56960a0f68c2bb63b31da15e53e96a4a9aa8b78fc5051e4d308ca9032210147faf3836b3f180a6","name":"proc_py310-linux_dict_sm@3.ldjson","type":"application/octet-stream","lastModified":1708211967079.248,"size":2916},{"id":"9cdd6643eb96b5db7bb9411dbca91220e727f31c7f80f148df8ec7c50010db92cb4a7334bbb52b886d3dca1e059606da33208b2346d361e4397db6bbeea638ea","href":"https://static.observableusercontent.com/files/9cdd6643eb96b5db7bb9411dbca91220e727f31c7f80f148df8ec7c50010db92cb4a7334bbb52b886d3dca1e059606da33208b2346d361e4397db6bbeea638ea","name":"proc_py310-linux_mdict_sm@3.ldjson","type":"application/octet-stream","lastModified":1708211967082.706,"size":2986},{"id":"f07c8cb6b670e26546eabd4c07db968e1428a110ec38c53a9440d9f1c284c5ec10b4031a8dd2157c2f8d1d92f65078706d54e4ff69d36aece19e53b6e620bdd8","href":"https://static.observableusercontent.com/files/f07c8cb6b670e26546eabd4c07db968e1428a110ec38c53a9440d9f1c284c5ec10b4031a8dd2157c2f8d1d92f65078706d54e4ff69d36aece19e53b6e620bdd8","name":"proc_py310-linux_mdict_lg@3.ldjson","type":"application/octet-stream","lastModified":1708211967163.918,"size":109846},{"id":"7131b5fdc7a93e0666a8eecbd64f3ef273f9c6b7a01aba04ea3a344b8e983de768e9b01676cda272f078875176255294a5ef2845c882f31b684cd4de055bf65c","href":"https://static.observableusercontent.com/files/7131b5fdc7a93e0666a8eecbd64f3ef273f9c6b7a01aba04ea3a344b8e983de768e9b01676cda272f078875176255294a5ef2845c882f31b684cd4de055bf65c","name":"proc_py310-linux_dict_lg@3.ldjson","type":"application/octet-stream","lastModified":1708211967175.002,"size":104104},{"id":"0207a7ec4c39d79a174f64bd5e4a1dfeeb17702f0218b102ea37f48a07980601cd378d53ade0b73cebf755edc379832d916f59ab4b05de1bfbd13c726edced0d","href":"https://static.observableusercontent.com/files/0207a7ec4c39d79a174f64bd5e4a1dfeeb17702f0218b102ea37f48a07980601cd378d53ade0b73cebf755edc379832d916f59ab4b05de1bfbd13c726edced0d","name":"proc_py310-linux_dict_sm@4.ldjson","type":"application/octet-stream","lastModified":1708212116303.512,"size":2920},{"id":"079c67c9fcb1b9f55ecf61054c0fe86ba283dd9af0dbfa0dc73518cffd1da44b4bbf892deff78d2d1b23d7e214b439cad67b755804960cb9a0664e1b5f4d4f85","href":"https://static.observableusercontent.com/files/079c67c9fcb1b9f55ecf61054c0fe86ba283dd9af0dbfa0dc73518cffd1da44b4bbf892deff78d2d1b23d7e214b439cad67b755804960cb9a0664e1b5f4d4f85","name":"proc_py310-linux_mdict_sm@4.ldjson","type":"application/octet-stream","lastModified":1708212116304.297,"size":2995},{"id":"926b3ea1b13f5aa031abb54b73d39b13b061a2ce649a2dd88c2cccc460f43113dc3fb972c540934fbb27141f3a999feff72b875e2ab85af7c9d9cfc948968c7b","href":"https://static.observableusercontent.com/files/926b3ea1b13f5aa031abb54b73d39b13b061a2ce649a2dd88c2cccc460f43113dc3fb972c540934fbb27141f3a999feff72b875e2ab85af7c9d9cfc948968c7b","name":"proc_py310-linux_mdict_lg@4.ldjson","type":"application/octet-stream","lastModified":1708212116401.72,"size":105852},{"id":"7508d7196367c0ab619a9ffd9546eef2150b558dfaf517e12e9fdf8c900135c0bbfc639d382ab2a136aa7e4489184cd4092095311d911daa39780d9cf729fe90","href":"https://static.observableusercontent.com/files/7508d7196367c0ab619a9ffd9546eef2150b558dfaf517e12e9fdf8c900135c0bbfc639d382ab2a136aa7e4489184cd4092095311d911daa39780d9cf729fe90","name":"proc_py310-linux_mdict_md@4.ldjson","type":"application/octet-stream","lastModified":1708212116309.919,"size":26501},{"id":"b13d8b91be68fd6b0cb4b0ab241c5c44710339bdb3e1f4d2ab7cd88e90c832e541354dbff1eaecd8df8f85519665bcef0e7ba5a0d2c68fa57d279c90b185b2ac","href":"https://static.observableusercontent.com/files/b13d8b91be68fd6b0cb4b0ab241c5c44710339bdb3e1f4d2ab7cd88e90c832e541354dbff1eaecd8df8f85519665bcef0e7ba5a0d2c68fa57d279c90b185b2ac","name":"proc_py310-linux_dict_md@4.ldjson","type":"application/octet-stream","lastModified":1708212116327.109,"size":24280},{"id":"980be54420a4c67d494f51d8e5ed50318914fd745161678752dd476555222f2928e380a5c864bd4a76414f78d2750cb46037a01c5d9c8e0265d7b7aea98720a5","href":"https://static.observableusercontent.com/files/980be54420a4c67d494f51d8e5ed50318914fd745161678752dd476555222f2928e380a5c864bd4a76414f78d2750cb46037a01c5d9c8e0265d7b7aea98720a5","name":"proc_py310-linux_dict_lg@4.ldjson","type":"application/octet-stream","lastModified":1708212116366.639,"size":101978},{"id":"f2fe53021d899255314fd4eec7e149b9f0140bceecee255a1179f723cb08b409b4ed6cde61d437bcad54f8aa110c97cf4d421d22359b4114b402651143ebb379","href":"https://static.observableusercontent.com/files/f2fe53021d899255314fd4eec7e149b9f0140bceecee255a1179f723cb08b409b4ed6cde61d437bcad54f8aa110c97cf4d421d22359b4114b402651143ebb379","name":"proc_py310-linux_dict_sm@5.ldjson","type":"application/octet-stream","lastModified":1708287296382.235,"size":2849},{"id":"fe97cb0920e4a375e4fd45cd2b9c2026d9d8d1028125de9120e419ff6f2b0f9020d0d708c70839ff3cbeabd95ba3c30dfdba46851c0031adb2216846e7acb155","href":"https://static.observableusercontent.com/files/fe97cb0920e4a375e4fd45cd2b9c2026d9d8d1028125de9120e419ff6f2b0f9020d0d708c70839ff3cbeabd95ba3c30dfdba46851c0031adb2216846e7acb155","name":"proc_py310-linux_mdict_sm@5.ldjson","type":"application/octet-stream","lastModified":1708287296397.659,"size":3603},{"id":"c60305423eff35a1201668438911101944a33582f0c18d3d6b7f1fd955e537c55e0e59137f5f88b8cf2a63191a989f75532faad2abb1ce19613341f248d7a948","href":"https://static.observableusercontent.com/files/c60305423eff35a1201668438911101944a33582f0c18d3d6b7f1fd955e537c55e0e59137f5f88b8cf2a63191a989f75532faad2abb1ce19613341f248d7a948","name":"proc_py310-linux_dict_md@5.ldjson","type":"application/octet-stream","lastModified":1708287296383.359,"size":24890},{"id":"70fa75a910bf0273f54391994bff9437ecec560393560bb1c46e769f4e46900b269750af21208bb059d208969569548e9de396da6d838f83678645e59c8806e2","href":"https://static.observableusercontent.com/files/70fa75a910bf0273f54391994bff9437ecec560393560bb1c46e769f4e46900b269750af21208bb059d208969569548e9de396da6d838f83678645e59c8806e2","name":"proc_py310-linux_mdict_md@5.ldjson","type":"application/octet-stream","lastModified":1708287296401.945,"size":26918},{"id":"a093a70c684602bf1a872a678bd3e97f82ceffe7101a624bb1c146de4c48a57527cc511386df3ac54a176eb8d22ce3762bae61bf5431ad774570dfde94fabe21","href":"https://static.observableusercontent.com/files/a093a70c684602bf1a872a678bd3e97f82ceffe7101a624bb1c146de4c48a57527cc511386df3ac54a176eb8d22ce3762bae61bf5431ad774570dfde94fabe21","name":"proc_py310-linux_mdict_lg@5.ldjson","type":"application/octet-stream","lastModified":1708287296482.854,"size":104791},{"id":"8b5fa23c6e35ac16bbde76ddb9a16c5eef499db2a4c6bfd18399570515040a420be3e27e80e01a01b741bdb1d94749132d40486f8e83d36acc02cb50392c156e","href":"https://static.observableusercontent.com/files/8b5fa23c6e35ac16bbde76ddb9a16c5eef499db2a4c6bfd18399570515040a420be3e27e80e01a01b741bdb1d94749132d40486f8e83d36acc02cb50392c156e","name":"proc_py310-linux_dict_lg@5.ldjson","type":"application/octet-stream","lastModified":1708287296463.797,"size":102282},{"id":"7131b5fdc7a93e0666a8eecbd64f3ef273f9c6b7a01aba04ea3a344b8e983de768e9b01676cda272f078875176255294a5ef2845c882f31b684cd4de055bf65c","href":"https://static.observableusercontent.com/files/7131b5fdc7a93e0666a8eecbd64f3ef273f9c6b7a01aba04ea3a344b8e983de768e9b01676cda272f078875176255294a5ef2845c882f31b684cd4de055bf65c","name":"proc_py310-linux_dict_lg@2.ldjson","type":"application/octet-stream","lastModified":1707114196362.218,"size":104104},{"id":"13017ed24987d40915c40d096a966a6c7478a1a2ee14f015bc56960a0f68c2bb63b31da15e53e96a4a9aa8b78fc5051e4d308ca9032210147faf3836b3f180a6","href":"https://static.observableusercontent.com/files/13017ed24987d40915c40d096a966a6c7478a1a2ee14f015bc56960a0f68c2bb63b31da15e53e96a4a9aa8b78fc5051e4d308ca9032210147faf3836b3f180a6","name":"proc_py310-linux_dict_sm@2.ldjson","type":"application/octet-stream","lastModified":1707114229534.119,"size":2916},{"id":"3bee4c9d53211f7856ce179232a9c3da8ed1ba8f3ad52fcfd84462e368309654c14130e907e7645e455e191e8dd89483844bcbf22dd15878e27ac0e94c742faa","href":"https://static.observableusercontent.com/files/3bee4c9d53211f7856ce179232a9c3da8ed1ba8f3ad52fcfd84462e368309654c14130e907e7645e455e191e8dd89483844bcbf22dd15878e27ac0e94c742faa","name":"proc_py310-linux_mdict_md@2.ldjson","type":"application/octet-stream","lastModified":1707114229555.113,"size":27401},{"id":"9cdd6643eb96b5db7bb9411dbca91220e727f31c7f80f148df8ec7c50010db92cb4a7334bbb52b886d3dca1e059606da33208b2346d361e4397db6bbeea638ea","href":"https://static.observableusercontent.com/files/9cdd6643eb96b5db7bb9411dbca91220e727f31c7f80f148df8ec7c50010db92cb4a7334bbb52b886d3dca1e059606da33208b2346d361e4397db6bbeea638ea","name":"proc_py310-linux_mdict_sm@2.ldjson","type":"application/octet-stream","lastModified":1707114229565.315,"size":2986},{"id":"c3186bf1052b6d435c1fa7fe0af55777f27c4740ea2880ea3038ac07e1afd897e20783335b95b5da2da243411fc15329e7bb598e0898a9c8bbeb4df143c7bcd9","href":"https://static.observableusercontent.com/files/c3186bf1052b6d435c1fa7fe0af55777f27c4740ea2880ea3038ac07e1afd897e20783335b95b5da2da243411fc15329e7bb598e0898a9c8bbeb4df143c7bcd9","name":"proc_py310-linux_dict_md@2.ldjson","type":"application/octet-stream","lastModified":1707114229580.971,"size":24691},{"id":"f07c8cb6b670e26546eabd4c07db968e1428a110ec38c53a9440d9f1c284c5ec10b4031a8dd2157c2f8d1d92f65078706d54e4ff69d36aece19e53b6e620bdd8","href":"https://static.observableusercontent.com/files/f07c8cb6b670e26546eabd4c07db968e1428a110ec38c53a9440d9f1c284c5ec10b4031a8dd2157c2f8d1d92f65078706d54e4ff69d36aece19e53b6e620bdd8","name":"proc_py310-linux_mdict_lg@2.ldjson","type":"application/octet-stream","lastModified":1707114229678.054,"size":109846}] 30:{"id":"0fd63725fc4d8f2d","slug":"pypocketmap-benchmarks","title":"pypocketmap benchmarks","thumbnail":"599ad2f4125fef7b65e629c79a2fef30615253358cd80e7699bfdb58d6b57225","access_level":"public","create_time":"$D2024-02-04T19:48:38.457Z","update_time":"$D2024-02-18T20:15:38.026Z","view_time":null,"first_public_time":"$D2024-02-04T19:48:38.457Z","has_unpublished_edits":null,"fork":true,"stars":0,"starred":false,"owner":{"type":"team","id":"1e53aaef20a0f828","login":"dylan-burati-ws","name":"Dylan Burati's Workspace","avatar_url":"https://avatars-next.observableusercontent.com/b71a672b142686bf953787d1747635f19e01fee36238acb451705e4e40dfaf14","tier":"starter_2024","bio":""},"version":325,"publish_version":null,"role":null,"authored":false,"authors":[{"type":"individual","id":"21003a4c65964963","login":"dylan-burati","name":"Dylan Burati","avatar_url":"https://avatars-next.observableusercontent.com/9e9b8dcddf739df6bc446adff1fe88371958b926887e70d71d97597dfdc40e73","tier":null,"bio":""}],"members":[{"type":"individual","id":"21003a4c65964963","login":"dylan-burati","name":"Dylan Burati","avatar_url":"https://avatars-next.observableusercontent.com/9e9b8dcddf739df6bc446adff1fe88371958b926887e70d71d97597dfdc40e73","tier":null,"role":"editor"},{"type":"team","id":"1e53aaef20a0f828","login":"dylan-burati-ws","name":"Dylan Burati's Workspace","avatar_url":"https://avatars-next.observableusercontent.com/b71a672b142686bf953787d1747635f19e01fee36238acb451705e4e40dfaf14","tier":"starter_2024","role":"editor"}],"comments":[],"custom_thumbnail":false,"trashed":false,"lapsed":null,"subscribed":null,"forks":0,"collection":null,"type":"notebook","body":{"titling":"auto","title":"pypocketmap benchmarks","cells":[{"id":0,"value":"# pypocketmap benchmarks\n\nThe code at https://github.com/dylanburati/pypocketmap/tree/main/tests/profile compares the performance of the builtin Python `dict` with a memory-efficient C implementation `pypocketmap[str, int64]` for a specific workload: counting occurrences of each unique word in a randomly generated stream 1, 8, or 32 million words long. These charts are generated from commit `80b7378f`, on a laptop with the following specs:\n\n```\nLinux 6.2.0-39-generic #40~22.04.1-Ubuntu\nArchitecture: x86_64\nCPU(s): 12\nModel name: Intel(R) Core(TM) i7-9750H CPU @ 2.60GHz\nMemory block size: 128M\nTotal online memory: 15.9G\n```","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":18,"value":"viewof radios = Inputs.radio(Object.keys(benchmarks), {label: \"Environment\", value: Object.keys(benchmarks)[0]})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":91,"value":"Plot.plot({\n title: \"Timings\",\n x: { label: \"Time to finish (s)\", inset: 5, domain: [0, Math.max(1, ...summaryProcessed.map(e => e.time))] },\n y: { label: \"Max observed memory (GiB)\", inset: 5, domain: [0, Math.max(1, ...summaryProcessed.map(e => e.memory))] },\n grid: summaryProcessed.length > 0,\n marks: [\n Plot.frame(),\n ...(summaryProcessed.length > 0 ? [] : [\n Plot.text([\"No data\"], { frameAnchor: \"middle\" })\n ]),\n Plot.dot(summaryProcessed, {\n x: \"time\",\n y: \"memory\",\n stroke: \"implementation\",\n channels: {implementation: \"implementation\", size: \"size\"},\n tip: true\n }),\n ]\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":4,"value":"Plot.plot({\n title: \"Memory during runs\",\n x: { label: \"Time elapsed (s)\", inset: 5 },\n y: { label: \"Resident Set Size (GiB)\", inset: 5, domain: [0, Math.max(1, ...benchmarksProcessed.map(e => e.mem))] },\n color: { legend: true },\n grid: true,\n fy: { label: null },\n marginRight: 50,\n width: 650,\n marks: [\n Plot.frame(),\n Plot.lineY(benchmarksProcessed.filter(d => d.size !== \"small\"), { x: \"timestamp\", y: \"mem\", fy: \"size\", stroke: \"implementation\", marker: \"circle\" })\n ]\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":234,"value":"$32","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":309,"value":"

","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":311,"value":"Using a `dict` to map to identifiers will use significantly more memory per item than this test does,\nbecause Python will [always use pre-allocated objects for the numbers -5 through 256](https://github.com/pablogsal/python-horror-show).","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":57,"value":"{\n const benchmarksSplit = [\n ...benchmarksProcessed.filter(e => e.size === \"large\").map((elt) => ({ ...elt, cpu: elt.mainCPU, whichThreads: \"Main Thread\" })),\n // ...benchmarksProcessed.map((elt) => ({ ...elt, cpu: elt.backCPU, whichThreads: \"Background Threads\" }))\n // .reduce((acc, cur) => [...acc, { ...cur, cpu: acc.length ? 0.5*acc[acc.length - 1].cpu+0.5*cur.cpu : cur.cpu }], []),\n ];\n return false || Plot.plot({\n title: \"CPU during runs (large input only)\",\n x: { label: \"Time elapsed (s)\", inset: 5 },\n y: { label: \"CPU (%)\", inset: 5 },\n grid: true,\n color: {legend: true},\n marks: [\n Plot.frame(),\n Plot.lineY(benchmarksSplit, {\n x: \"timestamp\", \n y: \"cpu\",\n stroke: \"implementation\",\n }),\n ]\n })\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":155,"value":"

","pinned":false,"hidden":false,"mode":"html","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"","database":"$undefined"},{"id":10,"value":"parseLdjson = async (stream) => {\n const decoder = new TextDecoderStream(\"utf-8\");\n const reader = stream.pipeThrough(decoder).getReader();\n let done, value;\n let tail = '';\n const result = [];\n while (({done, value} = await reader.read()), !done) {\n const lines = (tail + value).split('\\n');\n tail = lines.pop();\n result.push(...lines.map(txt => JSON.parse(txt)));\n }\n if (tail) result.push(JSON.parse(tail));\n return result;\n};","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":6,"value":"benchmarks = ({\n \"py310-linux\": {\n proc: [\n {\n implementation: \"dict\",\n small: await FileAttachment(\"proc_py310-linux_dict_sm@5.ldjson\").stream().then(parseLdjson),\n medium: await FileAttachment(\"proc_py310-linux_dict_md@5.ldjson\").stream().then(parseLdjson),\n large: await FileAttachment(\"proc_py310-linux_dict_lg@5.ldjson\").stream().then(parseLdjson),\n },\n {\n implementation: \"mdict\",\n small: await FileAttachment(\"proc_py310-linux_mdict_sm@5.ldjson\").stream().then(parseLdjson),\n medium: await FileAttachment(\"proc_py310-linux_mdict_md@5.ldjson\").stream().then(parseLdjson),\n large: await FileAttachment(\"proc_py310-linux_mdict_lg@5.ldjson\").stream().then(parseLdjson),\n },\n ]\n },\n});","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":21,"value":"$33","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":207,"value":"benchmarksProcessed = allProcessed[0]","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":87,"value":"summaryProcessed = allProcessed[1]","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"},{"id":238,"value":"sizes = {\n const lst = [...new Set(summaryProcessed.map((e) => parseInt(e.stdout.split('\\n')[1].split(/\\s+/)[1])))];\n lst.sort((a, b) => a - b);\n return lst;\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","providerMetadata":"$undefined","attachments":"$undefined","output":"$undefined","database":"$undefined"}],"files":"$W34","stdlib":"1"}} 31:"$30:body"