The throughput tables: the timed cells of a results file as rows, in groups of one language and source, each mode's cells led by their geometric mean.
8 declarations
throughput_matrix.ts view source
(results: Pick<MatrixResults, "cells">): boolean import {results_have_throughput} from '@ryanatkn/syntax-highlighter-bench/throughput_matrix.js'; Whether a results file holds any timed cell value.
resultsPick<MatrixResults, "cells">boolean throughput_matrix.ts view source
readonly { id: ThroughputUnit; label: string; }[] import {THROUGHPUT_UNITS} from '@ryanatkn/syntax-highlighter-bench/throughput_matrix.js'; Every throughput unit, with its label, in display order.
throughput_matrix.ts view source
ThroughputSelection import type {ThroughputSelection} from '@ryanatkn/syntax-highlighter-bench/throughput_matrix.js'; Which cells the throughput tables show, and in which unit.
unittype ThroughputUnit
throughput_matrix.ts view source
ThroughputUnit type "ns_per_op" | "ops_per_sec" | "mb_per_sec"
import type {ThroughputUnit} from '@ryanatkn/syntax-highlighter-bench/throughput_matrix.js'; The number a throughput cell is shown as. mb_per_sec is the one to compare
across inputs of different sizes.
throughput_matrix.ts view source
(value: { ns_per_op: number; ops_per_sec: number; mb_per_sec: number; p10_ns: number; p90_ns: number; spread: number; pass_medians_ns: number[]; iterations: number; rss_peak_bytes: number; attempts: number; }): MatrixBadge[] import {to_throughput_badges} from '@ryanatkn/syntax-highlighter-bench/throughput_matrix.js'; The marks on a value whose rounds disagreed or that was measured again.
valuethroughput_matrix.ts view source
(value: { ns_per_op: number; ops_per_sec: number; mb_per_sec: number; p10_ns: number; p90_ns: number; spread: number; pass_medians_ns: number[]; iterations: number; rss_peak_bytes: number; attempts: number; }): string[] import {to_throughput_details} from '@ryanatkn/syntax-highlighter-bench/throughput_matrix.js'; Everything a cell value records, as lines for the detail view.
valuestring[] throughput_matrix.ts view source
(results: MatrixResults, selection: ThroughputSelection, libraries: readonly string[]): MatrixGroup[] import {to_throughput_groups} from '@ryanatkn/syntax-highlighter-bench/throughput_matrix.js'; The selected cells as groups of one language and one source. Within a group each selected mode's cells follow a geometric mean over them, so the means read as the group's summary and the cells as what they summarize.
resultsselectionlibrariesthe selected library ids, which decide which cells are complete and so enter a mean
readonly string[]throughput_matrix.ts view source
(results: MatrixResults, selection: ThroughputSelection, libraries: readonly string[]): MatrixGroup import {to_throughput_summary} from '@ryanatkn/syntax-highlighter-bench/throughput_matrix.js'; The geometric means across every selected language: one row for each source and mode. Sources are never pooled, so a mean over home inputs always names the library whose samples they are.
resultsselectionlibrariesreadonly string[]MatrixGroup