29:null 2a:["$","title",null,{"children":"cSP v3 | Dominik's Workspace | Observable"}] 2f:T664,

${timeline({ periods: [ { onset: labelsNN.get(sample.id).on.median, offset: labelsNN.get(sample.id).off.median, y: [0, 1], type: `Pred: NN (${sample.split})` }, { onset: labelsNN.get(sample.id).on.min, offset: labelsNN.get(sample.id).on.max, y: [0, 1], type: `Pred: NN uncertainty` }, { onset: labelsNN.get(sample.id).off.min, offset: labelsNN.get(sample.id).off.max, y: [0, 1], type: `Pred: NN uncertainty` }, { onset: cspData.get(sample.id).csp_label[0], offset: cspData.get(sample.id).csp_label[1], y: [-1, 0], type: "Manual" } ], options: { subtitle: `cSP: NN=${( labelsNN.get(sample.id).duration.median * 1000 ).toFixed(1)}ms, manual=${( (sample.csp_label[1] - sample.csp_label[0]) * 1000 ).toFixed(1)}ms` } })}

${scatter({ channels: { x: (x) => x.manual * 1000, y: (y) => y.NN.median * 1000, stroke: "split", r: (x) => x.NN.std }, name: "NN", subtitle: `mean derivation: ${summary.duration.train.NN.err.toFixed(1)}/${summary.duration.test.NN.err.toFixed(1)}ms (train/test). Within ${scatterSettings.tol}ms: ${(summary.duration.train.NN.within*100).toFixed(1)}/${(summary.duration.test.NN.within*100).toFixed(1)}% (train/test). The size of the dots represents the uncertainty of the netowork prediction` })}

30:T4c0,

${timeline({ periods: [ { onset: labelsJulkunen.get(sample.id)[0], offset: labelsJulkunen.get(sample.id)[1], y: [0, 1], type: `Pred: Julkunen` }, { onset: cspData.get(sample.id).csp_label[0], offset: cspData.get(sample.id).csp_label[1], y: [-1, 0], type: "Manual" } ], options: { subtitle: `cSP: Julkunen=${( (labelsJulkunen.get(sample.id)[1] - labelsJulkunen.get(sample.id)[0]) * 1000 ).toFixed(1)}ms, manual=${( (sample.csp_label[1] - sample.csp_label[0]) * 1000 ).toFixed(1)}ms` } })}

${scatter({ channels: { x: (x) => x.manual * 1000, y: (y) => y.Julkunen.value * 1000, stroke: "black", }, name: "Julkunen", subtitle: `mean derivation: ${d3 .mean(cspTable, (x) => x.Julkunen.err * 1000) .toFixed(1)}ms. Within ${scatterSettings.tol}ms: ${( 100 * d3.mean(cspTable, (x) => x.Julkunen.err * 1000 <= scatterSettings.tol) ).toFixed(1)}%` })}

31:T44f,Plot.plot({ color: { legend: true }, y: { label: "%" }, x: { type: "log", label: "ms" }, title: "Percentage of test samples within a given accuracy", subtitle: "For a given error tolerance this shows how many test samples are classified within the given tolerance.", width: 1000, grid: true, marks: [ Plot.ruleX([scatterSettings.tol], { stroke: "gray" }), Plot.line( logspace(0, 2, 50).flatMap((x) => [ { type: "Julkunen", error: x, Within: d3.mean( cspTable.filter((x) => x.type == "duration" && x.split == "test"), (e) => e["Julkunen"].err <= x / 1000 ) }, { type: `NN`, error: x, Within: d3.mean( errorsNNtimeSeries.get(epoch).filter((x) => x.split == "test"), (e) => e.error <= x / 1000 ) } ]), { x: "error", y: (x) => 100 * x.Within, stroke: "type", tip: { format: { x: (s) => `${s.toFixed(1)}ms` } } } ), Plot.ruleY([0]) ] })32:T67d,

${timeline({ periods: [ { onset: labelsNNtimeSeries.get(sample.id).get("on").get(epoch).median, offset: labelsNNtimeSeries.get(sample.id).get("off").get(epoch).median, y: [0, 1], type: `Pred: NN (${sample.split})` }, { onset: cspData.get(sample.id).csp_label[0], offset: cspData.get(sample.id).csp_label[1], y: [-1, 0], type: "Manual" } ], options: { subtitle: `cSP: NN=${( labelsNN.get(sample.id).duration.median * 1000 ).toFixed(1)}ms, manual=${( (sample.csp_label[1] - sample.csp_label[0]) * 1000 ).toFixed(1)}ms` } })}

${Plot.plot({ title: "Training Dynamics", width: 600, height: 400, y: { type: "log" }, x: { type: "sqrt" }, marks: [ Plot.ruleY([0]), Plot.ruleX([epoch]), Plot.lineY( seriesNN.filter( (x) => x.type == "duration" && (x.split == "train" || x.split == "test") ), Plot.windowY({ x: "epoch", y: "error", z: "label", fy: "split", k: 10, reduce: "mean", opacity: (x) => (x.label == sample.id ? 1 : 0.1), stroke: (x) => (x.label == sample.id ? "red" : "black"), sort: (x) => x.label == sample.id, tip: { format: { stroke: false, opacity: false, y: (d) => (1000 * d).toFixed(1) + "ms" } } }) ) ] })}

33:T627,models = [ { //pred_log_all_data_xresnet1d18plus_512.json file: "pred_log_all_data_xresnet1d18plus_256.json", training: "all", model: "xresnet1d18plus", windowSize: 256 }, { //pred_log_all_data_xresnet1d18plus_512.json file: "pred_log_all_data_xresnet1d18plus_512.json", training: "all", model: "xresnet1d18plus", windowSize: 512 }, { //pred_log_all_data_xresnet1d34_deepplus_256.json file: "pred_log_all_data_xresnet1d34_deepplus_256.json", training: "all", model: "xresnet1d34_deepplus", windowSize: 256 }, { //pred_log_andre_data_xresnet1d18_deepplus_256.json file: "pred_log_andre_data_xresnet1d18_deepplus_256.json", training: "andre", model: "xresnet1d18_deepplus", windowSize: 256 }, { //pred_log_andre_data_xresnet1d18plus_256.json file: "pred_log_andre_data_xresnet1d18plus_256.json", training: "andre", model: "xresnet1d18plus", windowSize: 256 }, { //pred_log_katrin_data_xresnet1d18_deepplus_256.json file: "pred_log_katrin_data_xresnet1d18_deepplus_256.json", training: "katrin", model: "xresnet1d18_deepplus", windowSize: 256 }, { //pred_log_katrin_data_xresnet1d18plus_256.json file: "pred_log_katrin_data_xresnet1d18plus_256.json", training: "katrin", model: "xresnet1d18plus", windowSize: 256 }, { //pred_log_katrin_data_xresnet1d18plus_512.json file: "pred_log_katrin_data_xresnet1d18plus_512.json", training: "katrin", model: "xresnet1d18plus", windowSize: 512 } ]34:T4b2,scatter = ({ channels, subtitle, name }) => { const current_markers = scatterSettings.show == "Duration" ? ["duration"] : ["on", "off"]; return Plot.plot({ x: { label: "Manual cSP", domain: [0, 280] }, y: { label: name + " cSP", domain: [0, 280] }, aspectRatio: 1, width: 400, grid: true, title: "cSP", symbol: { legend: true }, color: { legend: true }, r: { legend: true }, subtitle: subtitle, marks: [ Plot.dot(cspTable, { fill: (x) => (x.id == sample.id ? "red" : "white"), opacity: (x) => (x.id == sample.id ? 1 : 0.3), filter: (x) => current_markers.includes(x["type"]), symbol: "type", channels: { sample: "id" }, ...channels, tip: { format: { sample: true, opacity: false, y: (d) => `${d.toFixed(1)}ms`, x: (d) => `${d.toFixed(1)}ms`, stroke: false, r: false, fill: false } } }), Plot.area([0.0, 300], { x1: (x) => x, y1: (x) => x - scatterSettings.tol, y2: (x) => x + scatterSettings.tol, opacity: 0.1 }) ] }); }35:T419,timeline = ({ periods, options, marks, id = sample.id }) => { return Plot.plot({ marks: [ Plot.frame({ strokeOpacity: 0.1 }), ...(timelineSettings.show.includes("Signal") ? [Plot.lineY(cspData.get(id).signal)] : []), ...(timelineSettings.show.includes("Derivative") ? [ Plot.lineY(derivative(cspData.get(id).signal), { stroke: "blue" }) ] : []), Plot.ruleX([0]), Plot.rect(periods, { x1: (x) => x.onset * frequency, x2: (x) => x.offset * frequency, y1: (x) => x.y[0], y2: (x) => x.y[1], fill: "type", opacity: 0.2 }) ], x: { type: "linear", domain: [0, timelineSettings.x], transform: (d) => d / frequency, label: "time (s)" }, y: { type: "linear", domain: [-0.005 / timelineSettings.y, 0.005 / timelineSettings.y] }, grid: true, color: { legend: true }, title: "Sample #" + id, ...options }); }36:[] 2d:{"id":"7239a061268131d9","slug":null,"title":"cSP v3","thumbnail":"b0df839482f543672bc46e8afbb84f3adf26826fdb8eedba8dda07813774eb28","access_level":"public","create_time":"$D2023-12-23T08:41:51.316Z","update_time":"$D2024-01-08T15:41:31.871Z","view_time":null,"first_public_time":"$D2023-12-23T08:41:51.316Z","fork":{"id":"61ce68d174bd19cf","slug":null,"title":"cSP v2","owner":{"id":"6fbae63626342e49","login":"dominik2-ws","name":"Dominik's Workspace"}},"stars":0,"starred":false,"owner":{"type":"team","id":"6fbae63626342e49","login":"dominik2-ws","name":"Dominik's Workspace","avatar_url":"https://avatars-next.observableusercontent.com/29d54f7b0ada4485ab66a5b753687a16ebd840cee3486735bc72600f3928a351","tier":"starter_2024"},"version":2247,"publish_version":null,"role":null,"authored":false,"authors":[{"type":"individual","id":"8d3a001f2b495452","login":"dominik2","name":"Dominik","avatar_url":"https://avatars-next.observableusercontent.com/2be6a595e5ee20eb59db0ea25f43e8e6d39f847748b68407f16331b31d5d6d82","tier":null}],"members":[{"type":"individual","id":"8d3a001f2b495452","login":"dominik2","name":"Dominik","avatar_url":"https://avatars-next.observableusercontent.com/2be6a595e5ee20eb59db0ea25f43e8e6d39f847748b68407f16331b31d5d6d82","tier":null,"role":"editor"},{"type":"team","id":"6fbae63626342e49","login":"dominik2-ws","name":"Dominik's Workspace","avatar_url":"https://avatars-next.observableusercontent.com/29d54f7b0ada4485ab66a5b753687a16ebd840cee3486735bc72600f3928a351","tier":"starter_2024","role":"editor"}],"comments":[],"custom_thumbnail":false,"trashed":false,"subscribed":null,"forks":0,"collection":null,"type":"notebook","body":{"title":"cSP v3","cells":[{"id":0,"value":"# cSP v3","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":1136,"value":"Choose the training data set\n\n- Katrin: 80% training, 20% testing (André: 100% testing)\n- André: 80% training, 20% testing (Katrin: 100% testing)\n- All: 80% training, 20% testing","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":1838,"value":"viewof modelData = Inputs.select(Array.from(modelsByData.entries()), {\n label: \"Training Data\",\n format: (x) => x[0]\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1842,"value":"viewof model = Inputs.select(modelData[1], {\n format: (x) => `${x.model}:${x.windowSize}`,\n label: \"Training Model\"\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1134,"value":"---","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":71,"value":"viewof sample = Inputs.select(Array.from(cspData.values()), {\n label: \"Sample ID\",\n format: (x) => x.id\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":612,"value":"viewof timelineSettings = Inputs.form({\n x: Inputs.range([0.2, 1], { step: 0.05, label: \"x-Range (s)\", value: 0.5 }),\n y: Inputs.range([0.5, 100], { step: 0.1, label: \"y-Scale\", value: 2 }),\n show: Inputs.checkbox([\"Signal\", \"Derivative\"], {\n label: \"Plot\",\n value: [\"Signal\", \"Derivative\"]\n })\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":611,"value":"viewof scatterSettings = Inputs.form({\n tol: Inputs.range([1, 20], {\n step: 1,\n label: \"Duration tolerance (ms)\",\n value: 10\n }),\n show: Inputs.radio([\"On/Off\", \"Duration\"], {\n label: \"Plot\",\n value: \"Duration\"\n })\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":492,"value":"$2f","pinned":false,"hidden":false,"mode":"html","author":"$undefined","summary":"$undefined"},{"id":327,"value":"viewof plot_julkunen_threshold = Inputs.range([0.000001, 0.0001], {\n step: 0.000001,\n label: \"Julkunen threshold\",\n value: 0.000041\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":367,"value":"$30","pinned":false,"hidden":false,"mode":"html","author":"$undefined","summary":"$undefined"},{"id":1027,"value":"Inputs.table(cspLabelsCombined, {\n columns: [\n \"id\",\n \"split\",\n \"csp_label\",\n \"NN_median\",\n \"NN_std\",\n \"NN_error\",\n \"Julkunen\",\n \"Julkunen_error\"\n ],\n header: {\n id: \"Sample\",\n split: \"Split\",\n csp_label: \"Label\",\n NN_median: \"NN\",\n NN_std: \"±\",\n NN_error: \"Err\",\n Julkunen_error: \"Err\"\n },\n format: {\n NN_median: durationFormat,\n csp_label: durationFormat,\n Julkunen: durationFormat,\n NN_error: (x) => (x * 1000).toFixed(1),\n Julkunen_error: (x) => (x * 1000).toFixed(1),\n NN_std: (x) => (x * 1000).toFixed(1),\n Error: (x) => (x * 1000).toFixed(1)\n },\n width: {\n id: 80,\n split: 50,\n NN_std: 30,\n NN_median: 160,\n Julkunen: 160,\n csp_label: 160,\n NN_error: 40,\n Julkunen_error: 40\n },\n sort: {\n NN_median: (x) => x[1] - x[0],\n csp_label: (x) => x[1] - x[0],\n Julkunen: (x) => x[1] - x[0]\n },\n multiple: false,\n required: false\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1182,"value":"## Summary","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":2208,"value":"The following table shows the performance of the algorithm for test/training samples separately. The neural network (on the test set) achieves a *mean error* of **${summary.duration.test.NN.err.toFixed(1)}ms** and classifies **${(summary.duration.test.NN.within*100).toFixed(1)}%** within a tolerance of **${scatterSettings.tol}ms**.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":1724,"value":"{\n const models = Object.keys(summary.duration.test).filter((x) => x != \"num\");\n return Inputs.table(\n Object.entries(summary.duration).flatMap(([split, entry]) => ({\n split,\n num: entry.num,\n ...Object.fromEntries(\n Object.entries(entry).flatMap(([k, v]) =>\n typeof v == \"object\"\n ? Object.entries(v).map(([k2, v2]) => [k + \"_\" + k2, v2])\n : [[k, v]]\n )\n )\n })),\n {\n header: {\n split: \"Split\",\n num: \"#Samples\",\n ...Object.fromEntries(\n models.flatMap((id) => [\n [id + \"_err\", id + \" mean error\"],\n [id + \"_within\", `${id} within ${scatterSettings.tol}ms`]\n ])\n )\n },\n format: {\n num: (x) => x,\n ...Object.fromEntries(\n models.flatMap((id) => [\n [id + \"_err\", (v) => `${v.toFixed(1)}ms`],\n [id + \"_within\", (v) => `${(v * 100).toFixed(0)}%`]\n ])\n )\n }\n }\n );\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":2220,"value":"The distribution of the errors is displayed in the following plots. ","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":1274,"value":"bar(\"NN\")","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1343,"value":"bar(\"Julkunen\")","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":2017,"value":"$31","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":2060,"value":"### Worst test samples","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":2103,"value":"The 9 *worst* test samples are plotted below. The test samples classified outside the tolerance of **${scatterSettings.tol}ms** are *${cspLabelsCombined.filter(x=>x.NN_error>scatterSettings.tol/1000&&x.split==\"test\").sort((a,b)=>b.NN_error-a.NN_error).map(x=>x.id).join(\", \")}*","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":2100,"value":"Inputs.bind(\n Inputs.form({\n x: Inputs.range([0.2, 1], { step: 0.05, label: \"x-Range (s)\", value: 0.5 }),\n y: Inputs.range([0.5, 100], { step: 0.1, label: \"y-Scale\", value: 2 }),\n show: Inputs.checkbox([\"Signal\", \"Derivative\"], {\n label: \"Plot\",\n value: [\"Signal\", \"Derivative\"]\n })\n }),\n viewof timelineSettings\n)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":2077,"value":"

\n${cspTable\n .filter((x) => x.split == \"test\" && x.type == \"duration\")\n .sort((a, b) => -a.NN.err + b.NN.err)\n .slice(0, 9)\n .map((x) => NNtimeline(x.id))}\n

","pinned":false,"hidden":false,"mode":"html","author":"$undefined","summary":"$undefined"},{"id":1792,"value":"## NN Training Dynamics","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":2108,"value":"Inputs.bind(\n Inputs.select(Array.from(cspData.values()), {\n label: \"Sample ID\",\n format: (x) => x.id\n }),\n viewof sample\n)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":2194,"value":"viewof epoch = Inputs.range(d3.extent(epochs), {\n step: epochs[1] - epochs[0],\n value: epochs[epochs.length - 1],\n label: \"Epoch\"\n})","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1931,"value":"$32","pinned":false,"hidden":false,"mode":"html","author":"$undefined","summary":"$undefined"},{"id":132,"value":"---\n# Appendix","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":68,"value":"frequency = 3000","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":607,"value":"cspData = {\n const request = await fetch(\n `https://wirhabenzeit.github.io/csp/${model.training}_data.json`\n );\n var data = await request.json();\n var collator = new Intl.Collator(undefined, {\n numeric: true,\n sensitivity: \"base\"\n });\n data = data.sort((a, b) => collator.compare(a.id, b.id));\n return new Map(data.map((entry) => [entry.id, entry]));\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":605,"value":"## Julkunen Prediction","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":259,"value":"derivative = (signal) =>\n signal.slice(1, signal.length).map((x, i) => x - signal[i])","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":263,"value":"julkunenPrediction = (\n threshold = 0.0225,\n freq = 3000,\n minLen = 0.03,\n maxTime = 0.3\n ) =>\n (signal) => {\n const diff = derivative(signal).map((x) => Math.abs(x * freq * 1e-3));\n const minInd = Math.round(minLen * freq);\n const windowsOfSmallDerivative = [];\n var [i, j] = [0, 0];\n while (j < Math.min(diff.length, maxTime * freq)) {\n if (Math.abs(diff[j]) >= threshold) {\n if (j - i > minInd) windowsOfSmallDerivative.push([i, j - 1]);\n i = j + 1;\n j = i;\n } else j += 1;\n }\n if (j - i > minInd) windowsOfSmallDerivative.push([i, j - 1]);\n if (windowsOfSmallDerivative.length == 0) return [0, 0];\n return windowsOfSmallDerivative\n .sort((a, b) => d3.descending(a[1] - a[0], b[1] - b[0]))[0]\n .map((x) => x / freq);\n }","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":395,"value":"labelsJulkunen = new Map(\n Array.from(cspData, ([key, entry]) => [\n key,\n julkunenPrediction(plot_julkunen_threshold)(entry.signal)\n ])\n)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":602,"value":"## Neural Network Prediction","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":1827,"value":"$33","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1830,"value":"modelsByData = d3.group(models, (x) => x.training)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1447,"value":"seriesNN = {\n const request = await fetch(\n `https://wirhabenzeit.github.io/csp/${model.file}`\n );\n var data = await request.json();\n data.forEach((x) => {\n const label = cspData.get(x.label).csp_label;\n x.manual =\n x.type == \"on\"\n ? label[0]\n : x.type == \"off\"\n ? label[1]\n : label[1] - label[0];\n x.error = Math.abs(x.median - x.manual);\n x.split = cspData.get(x.label).split;\n });\n return data;\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1822,"value":"labelsNNtimeSeries = d3.rollup(\n seriesNN,\n (v) => v[0],\n (x) => x.label,\n (x) => x.type,\n (x) => x.epoch\n)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":2142,"value":"epochs = Array.from(\n Array.from(labelsNNtimeSeries.entries())[0][1].get(\"duration\").keys()\n)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":2136,"value":"errorsNNtimeSeries = new Map(\n epochs.map((epoch) => [\n epoch,\n d3.map(labelsNNtimeSeries, ([id, ts]) => ({\n id,\n split: cspData.get(id).split,\n error: ts.get(\"duration\").get(epoch).error\n }))\n ])\n)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1476,"value":"labelsNN = {\n const last_epoch = d3.max(seriesNN, (x) => x.epoch);\n return new d3.InternMap(\n d3\n .rollups(\n seriesNN.filter((x) => x.epoch == last_epoch),\n ([{ mean, median, std }]) => ({\n mean,\n median,\n std,\n min: mean - std,\n max: mean + std\n }),\n (x) => x.label,\n (x) => x.type\n )\n .map(([key, arrs]) => [key, Object.fromEntries(arrs)])\n );\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":595,"value":"## Data with predicitons ","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":1085,"value":"cspTable = Array.from(cspData.entries())\n .flatMap(([key, { id, split, csp_label }]) =>\n [\"on\", \"off\", \"duration\"].map((type) => ({\n id,\n split,\n type: type,\n manual: onOffObj(csp_label)[type],\n Julkunen: { value: onOffObj(labelsJulkunen.get(id))[type] },\n NN: { ...labelsNN.get(id)[type], value: labelsNN.get(id)[type].median }\n }))\n )\n .map((entry) => ({\n ...entry,\n NN: { ...entry.NN, err: Math.abs(entry.NN.value - entry.manual) },\n Julkunen: {\n ...entry.Julkunen,\n err: Math.abs(entry.Julkunen.value - entry.manual)\n }\n }))","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1174,"value":"cspLabelsCombined = Array.from(\n cspData.entries(),\n ([key, { csp_label, id, split }]) => ({\n id,\n split,\n csp_label,\n NN_median: [labelsNN.get(id).on.median, labelsNN.get(id).off.median],\n NN_std: labelsNN.get(id).duration.std,\n Julkunen: labelsJulkunen.get(id),\n NN_error: Math.abs(\n labelsNN.get(id).off.median -\n labelsNN.get(id).on.median -\n csp_label[1] +\n csp_label[0]\n ),\n Julkunen_error: Math.abs(\n labelsJulkunen.get(id)[1] -\n labelsJulkunen.get(id)[0] -\n csp_label[1] +\n csp_label[0]\n )\n })\n)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1219,"value":"summary = Object.fromEntries(\n d3\n .rollups(\n cspTable,\n (v) => ({\n ...Object.fromEntries(\n [\"NN\", \"Julkunen\"].map((name) => [\n name,\n {\n err: d3.mean(v, (e) => e[name].err * 1000),\n within: d3.mean(v, (e) =>\n e[name].err * 1000 <= scatterSettings.tol ? 1 : 0\n )\n }\n ])\n ),\n num: v.length\n }),\n (e) => e.type,\n (e) => e.split\n )\n .map(([key, entry]) => [key, Object.fromEntries(entry)])\n)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1502,"value":"onOffObj = (v) => ({ on: v[0], off: v[1], duration: v[1] - v[0] })","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1038,"value":"durationFormat = (x) =>\n `${(x[0] * 1000).toFixed(1)}-${(x[1] * 1000).toFixed(1)}ms (${(\n (x[1] - x[0]) *\n 1000\n ).toFixed(1)}ms)`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":592,"value":"## Plots","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined"},{"id":1419,"value":"function set(input, value) {\n input.value = value;\n input.dispatchEvent(new Event(\"input\", { bubbles: true }));\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1535,"value":"$34","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1807,"value":"$35","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":2063,"value":"NNtimeline = (id) =>\n timeline({\n periods: [\n {\n onset: labelsNN.get(id).on.median,\n offset: labelsNN.get(id).off.median,\n y: [0, 1],\n type: `Pred: NN`\n },\n {\n onset: labelsNN.get(id).on.min,\n offset: labelsNN.get(id).on.max,\n y: [0, 1],\n type: `Pred: NN uncertainty`\n },\n {\n onset: labelsNN.get(id).off.min,\n offset: labelsNN.get(id).off.max,\n y: [0, 1],\n type: `Pred: NN uncertainty`\n },\n {\n onset: cspData.get(id).csp_label[0],\n offset: cspData.get(id).csp_label[1],\n y: [-1, 0],\n type: \"Manual\"\n }\n ],\n options: {\n subtitle: `cSP: NN=${(labelsNN.get(id).duration.median * 1000).toFixed(\n 1\n )}ms, manual=${(\n (cspData.get(id).csp_label[1] - cspData.get(id).csp_label[0]) *\n 1000\n ).toFixed(1)}ms`,\n width: 300,\n height: 200\n },\n id\n })","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":1351,"value":"bar = (name) => {\n return Plot.plot({\n title: name,\n subtitle:\n \"Distribution of errors, separated by split (training/testing) and measurement type (onset/offset/duration)\",\n width: 1200,\n marginLeft: 80,\n marginRight: 80,\n x: { domain: [0, 150] },\n marks: [\n Plot.barX([{ split: \"exclude\" }, { split: \"test\" }, { split: \"train\" }], {\n x1: 0,\n x2: scatterSettings.tol,\n opacity: 0.1,\n y: \"split\"\n }),\n Plot.tickX(cspTable, {\n x: (x) => x[name].err * 1000,\n y: \"split\",\n opacity: (e) => (e.id == sample.id ? 1 : 0.2),\n stroke: (e) => (e.id == sample.id ? \"red\" : \"blue\"),\n fy: \"type\",\n channels: { title: (x) => x.id },\n tip: true\n })\n ]\n });\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"},{"id":2117,"value":"function logspace(a, b, len) {\n var arr, end, tmp, d;\n\n if (typeof a !== \"number\" || a !== a) {\n throw new TypeError(\n \"logspace()::invalid input argument. Exponent of start value must be numeric.\"\n );\n }\n if (typeof b !== \"number\" || b !== b) {\n throw new TypeError(\n \"logspace()::invalid input argument. Exponent of stop value must be numeric.\"\n );\n }\n if (arguments.length < 3) {\n len = 10;\n } else {\n if (len === 0) {\n return [];\n }\n }\n // Calculate the increment:\n end = len - 1;\n d = (b - a) / end;\n\n // Build the output array...\n arr = new Array(len);\n tmp = a;\n arr[0] = Math.pow(10, tmp);\n for (var i = 1; i < end; i++) {\n tmp += d;\n arr[i] = Math.pow(10, tmp);\n }\n arr[end] = Math.pow(10, b);\n return arr;\n} // end FUNCTION logspace()","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined"}],"files":"$W36","stdlib":"1","titling":"auto"}} 2e:"$2d:body"