32:T563, compute = function(fDec, fRec, decFirst = true) { const curstate=[]; const decArr=new Array(); const recArr=new Array(); patients.forEach(p => { for (let i = 0; i < +p["state.in_hospital.in"]; i++) { curstate.push({day_in: p.day}); } const today = p["day"]; const dec = p["state.deceased.hospital"]; const hout = p["state.in_hospital.out"]; if (decFirst) { for (let i = 0; i < dec; i++) { const timeToDeath = today - fDec(curstate, today).day_in; decArr[timeToDeath] = (~~decArr[timeToDeath])+1; } } for (let i = 0; i < hout; i++) { const timeToRelease = today - fRec(curstate, today).day_in; recArr[timeToRelease] = (~~recArr[timeToRelease])+1; } if (!decFirst) { for (let i = 0; i < dec; i++) { const timeToDeath = today - fDec(curstate, today).day_in; decArr[timeToDeath] = (~~decArr[timeToDeath])+1; } } }); const sumDec = decArr.reduce((res,a) => res + a); const sumRec = recArr.reduce((res,a) => res + a); const sumCur = curstate.map(s=>1).reduce((a,b) => a + b); return { "decArr": decArr, "sumDec": sumDec, "cumDec": dist2cum(decArr).map(a => a/sumDec), "recArr": recArr, "sumRec": sumRec, "cumRec": dist2cum(recArr).map(a => a/sumRec), "curstate": curstate, "sumCur": sumCur } }33:T50c,chart = function(arr, nDays=45) { const height = 300; const margin = ({top: 30, right: 0, bottom: 30, left: 40}); const x = d3.scaleBand() .domain(d3.range(nDays)) .range([margin.left, width - margin.right]) .padding(0.1); const xAxis = g => g .attr("transform", `translate(0,${height - margin.bottom})`) .call(d3.axisBottom(x).tickFormat(i => i).tickSizeOuter(0)); const y = d3.scaleLinear() .domain([0, d3.max(arr, d => d)]).nice() .range([height - margin.bottom, margin.top]); const yAxis = g => g .attr("transform", `translate(${margin.left},0)`) .call(d3.axisLeft(y).ticks(null, '')) .call(g => g.select(".domain").remove()) .call(g => g.append("text") .attr("x", -margin.left) .attr("y", 10) .attr("fill", "currentColor") .attr("text-anchor", "start") ); const svg = d3.create("svg") .attr("viewBox", [0, 0, width, height]); svg.append("g") .attr("fill", "blue") .selectAll("rect") .data(arr) .join("rect") .attr("x", (d, i) => x(i)) .attr("y", d => y(d)) .attr("height", d => y(0) - y(d)) .attr("width", x.bandwidth()); svg.append("g") .call(xAxis); svg.append("g") .call(yAxis); return svg.node(); }34:Td12,lineChartCI = function(data, max, color = "blue") { const height = 380; const yTickSize = 40; const yLabelOffset = -yTickSize + 35; const margin = ({top: 20, right: 0, bottom: 60, left: yTickSize}); const len = max || data.map(a => (a.d80 > 0)).lastIndexOf(true); const x = d3.scaleLinear() .domain([0,len]) .range([margin.left, width - margin.right]); const xAxis = g => g .attr("transform", `translate(0,${height - margin.bottom})`) .call(d3.axisBottom(x).ticks(len).tickFormat(x => (x%7==0) ? x : "").tickSizeOuter(0)) .call(g => g.selectAll(".tick line").clone() // grid lines .attr("stroke-opacity", 0.05) .attr("y1", -height)); const y = d3.scaleLinear() .domain([0, d3.max(data, d => d.d975)]).nice() .range([height - margin.bottom, margin.top]); const yAxis = g => g .attr("transform", `translate(${margin.left + 1},0)`) .call(d3.axisLeft(y).ticks(height / 100).tickSize(yTickSize)) // left axis .attr("stroke-opacity", 0.05) .call(g => g.selectAll(".tick line").clone() // grid lines .attr("stroke-opacity", 0.05) .attr("x2", width - margin.left - margin.right)) .call(g => g.select(".tick:first-of-type") // first tick opacity .attr("stroke-opacity", 1)) .call(g => g.selectAll(".tick text") // tick labels .style("text-anchor", "start") .attr("x", -yTickSize) .attr("y", -9) .attr("class", "xsmall-copy")) .call(g => g.select(".domain").remove()) // remove domain line .call(g => g.select(".tick:last-of-type text").clone() // unit label .attr("x", `${yLabelOffset}`) .attr("text-anchor", "start") .attr("class", "xsmall-copy font-bold") .text(data.y)); const lineMedian = d3.line() .x(d => x(d.days)) .y(d => y(d.median)); const lineMean = d3.line() .defined(d => !isNaN(d.mean)) .x(d => x(d.days)) .y(d => y(d.mean)); const svg = d3.create("svg") .attr("viewBox", [0, 0, width, height]) .attr("id", "chart"); svg.append("g") .call(xAxis); svg.append("g") .call(yAxis); svg.append("path") // 95% confidence interval .datum(data) .attr("fill", color) .attr("stroke", "none") .attr("opacity", "10%") .attr("stroke-width", 1.5) .attr("d", d3.area() .x(d => x(d.days)) .y0(d => y(d.d025)) .y1(d => y(d.d975))); svg.append("path") // 60% confidence interval .datum(data) .attr("fill", color) .attr("stroke", "none") .attr("opacity", "10%") .attr("stroke-width", 1.5) .attr("d", d3.area() .x(d => x(d.days)) .y0(d => y(d.d20)) .y1(d => y(d.d80))); svg.append("path") // line .datum(data) .attr("fill", "none") .attr("stroke-width", 1.5) .attr("stroke-linejoin", "round") .attr("stroke-linecap", "round") .attr("stroke", color) .attr("opacity", "100%") .attr("d", lineMedian); svg.append("path") // line .datum(data) .attr("fill", "none") .attr("stroke", color) .attr("stroke-dasharray", "10,10") .attr("stroke-width", 1.5) .attr("stroke-linejoin", "round") .attr("stroke-linecap", "round") .attr("d", lineMean); return svg.node(); }35:[] 30:{"title":"Sledilnik.org — Ocena trajanja hospitalizacij v Sloveniji","cells":[{"id":0,"value":"md`# Sledilnik.org — Ocena trajanja hospitalizacij v Sloveniji`","pinned":false,"hidden":false,"mode":"ojs"},{"id":6,"value":"patients_raw = d3.csv(\"https://raw.githubusercontent.com/sledilnik/data/master/csv/patients.csv\", d3.autoType);","pinned":false,"hidden":false,"mode":"ojs"},{"id":708,"value":"patients = {\n const arr = patients_raw.map(p => {\n p[\"state.in_hospital.in\"] = ~~p[\"state.in_hospital.in\"];\n p[\"state.in_hospital.out\"] = ~~p[\"state.in_hospital.out\"];\n p[\"state.icu.in\"] = ~~p[\"state.icu.in\"];\n p[\"state.icu.out\"] = ~~p[\"state.icu.out\"];\n p[\"state.deceased.hospital\"] = ~~p[\"state.deceased.hospital\"];\n p[\"state.deceased.hospital.icu\"] = ~~p[\"state.deceased.hospital.icu\"];\n if (p.day < 11) {\n p = {...p, \"state.in_hospital.out\": 0}; // Remove 19 recoveries (at the beginning, mild cases were put into H)\n }\n return p;\n });\n //Add 7 hospitalizations before 2020-03-11 (non-mild cases added at the beginning, not recorded in raw dataset)\n arr[2][\"state.in_hospital.in\"] = 1;\n arr[3][\"state.in_hospital.in\"] = 2;\n arr[4][\"state.in_hospital.in\"] = 1;\n arr[5][\"state.in_hospital.in\"] = 3;\n return arr;\n}","pinned":true,"hidden":false,"mode":"ojs"},{"id":612,"value":"md`## Iterative Algorithm\n\nUses CDF of the duration from the previous iteration to probabilistically remove active patients from the hospital. Converges to a smooth distribution centered around 11, regardless of the starting point (fifo, lifo or random).`","pinned":false,"hidden":false,"mode":"ojs"},{"id":636,"value":"md`### Statistics over ${iter_opts.n_runs} runs of ${iter_opts.n_iter} iterations`","pinned":false,"hidden":false,"mode":"ojs"},{"id":855,"value":"iter_opts = ({n_runs: 20, n_iter: 20});","pinned":false,"hidden":false,"mode":"ojs"},{"id":222,"value":"data_stat = {\n const arrRec = new Array(100).fill(null).map(() => []);\n const arrDec = new Array(100).fill(null).map(() => []);\n for (let t=0; t < iter_opts.n_runs; t++) {\n let newData = computeIter(data_rnd, iter_opts.n_iter);\n arrRec.forEach((a, idx) => {\n a.push(~~newData.recArr[idx]);\n // return a;\n });\n arrDec.forEach((a, idx) => {\n a.push(~~newData.decArr[idx]);\n })\n }\n \n //compute stats\n \n return {\n rec: arrRec.map(computeStats),\n dec: arrDec.map(computeStats)\n };\n}","pinned":false,"hidden":false,"mode":"ojs"},{"id":520,"value":"lineChartCI(data_stat.rec, 43)","pinned":false,"hidden":false,"mode":"ojs"},{"id":590,"value":"lineChartCI(data_stat.dec, 43, \"black\")","pinned":false,"hidden":false,"mode":"ojs"},{"id":821,"value":"md`### Aggregation of CDF Values`","pinned":false,"hidden":false,"mode":"ojs"},{"id":814,"value":"cdf_stat = {\n const arrRec = new Array(100).fill(null).map(() => []);\n const arrDec = new Array(100).fill(null).map(() => []);\n for (let t=0; t < iter_opts.n_runs; t++) {\n let newData = computeIter(data_rnd, iter_opts.n_iter);\n arrRec.forEach((a, idx) => {\n a.push((newData.cumRec[idx] === undefined) ? (idx > 20 ? 1 : 0) : newData.cumRec[idx]);\n // return a;\n });\n arrDec.forEach((a, idx) => {\n a.push((newData.cumDec[idx] === undefined) ? (idx > 20 ? 1 : 0) : newData.cumDec[idx]);\n })\n }\n \n //compute stats\n \n return {\n rec: arrRec.map(computeStats),\n dec: arrDec.map(computeStats)\n };\n}","pinned":false,"hidden":false,"mode":"ojs"},{"id":816,"value":"lineChartCI(cdf_stat.rec, 35)","pinned":true,"hidden":false,"mode":"ojs"},{"id":847,"value":"lineChartCI(cdf_stat.dec, 35)","pinned":true,"hidden":false,"mode":"ojs"},{"id":461,"value":"data_it = computeIter(data_rnd, 100)","pinned":false,"hidden":false,"mode":"ojs"},{"id":616,"value":"md`Chart: CDF of duration for deceased (after 100 iterations)`","pinned":false,"hidden":false,"mode":"ojs"},{"id":217,"value":"chart(data_it.cumDec, 35)","pinned":false,"hidden":false,"mode":"ojs"},{"id":620,"value":"md`Chart: CDF of duration of recovered (after 100 iterations)`","pinned":false,"hidden":false,"mode":"ojs"},{"id":235,"value":"chart(data_it.cumRec, 35)","pinned":false,"hidden":false,"mode":"ojs"},{"id":627,"value":"md`Resulting PDF after 50 iterations`","pinned":false,"hidden":false,"mode":"ojs"},{"id":428,"value":"bothCharts(data_it, 50)","pinned":true,"hidden":false,"mode":"ojs"},{"id":73,"value":"md`## Izračun trajanja hospitalizacij za umrle in odpuščene primere ob predpostavki FIFO\nAlgoritem iterira preko podatkov o primerih, sprejetih v bolnišnico, in vodi evidenco aktivnih oseb v bolnišnici po razdelkih glede na dan, ko so bili sprejeti. Dnevno umrle in odpuščene paciente upošteva tako, da zmanjša število aktivnih oseb v najbolj zgodnjem razdelku - first in, first out (FIFO). Umrle osebe se upoštevajo najprej, nato še odpuščene (ko ugibamo kateri od aktivnih bolnikov je umrl in kateri je bil odpuščen, za smrtni izid določimo tiste, ki so dlje v bolnišnici).\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":10,"value":"data_fifo = compute(st => st.shift(), st => st.shift(), true)","pinned":false,"hidden":false,"mode":"ojs"},{"id":75,"value":"md`## Porazdelitev trajanja bolnišničnega zdravljenja - ozdravljeni primeri`","pinned":false,"hidden":false,"mode":"ojs"},{"id":62,"value":"chart(data_fifo.recArr)","pinned":false,"hidden":false,"mode":"ojs"},{"id":99,"value":"md`## Porazdelitev trajanja bolnišničnega zdravljenja - smrtni primeri`","pinned":false,"hidden":false,"mode":"ojs"},{"id":65,"value":"chart(data_fifo.decArr)","pinned":false,"hidden":false,"mode":"ojs"},{"id":107,"value":"md`## Random instead of FIFO`","pinned":false,"hidden":false,"mode":"ojs"},{"id":118,"value":"data_rnd = compute(st => st.length == 0 ? null : st.splice(Math.floor(Math.random()*st.length),1)[0],\n st => st.length == 0 ? null : st.splice(Math.floor(Math.random()*st.length),1)[0],\n false);","pinned":false,"hidden":false,"mode":"ojs"},{"id":185,"value":"md`### Recovered (using random)`","pinned":false,"hidden":false,"mode":"ojs"},{"id":137,"value":"chart(data_rnd.recArr)","pinned":false,"hidden":false,"mode":"ojs"},{"id":189,"value":"md`### Deceased (using random)`","pinned":false,"hidden":false,"mode":"ojs"},{"id":130,"value":"chart(data_rnd.decArr)","pinned":true,"hidden":false,"mode":"ojs"},{"id":141,"value":"md`## LIFO`","pinned":false,"hidden":false,"mode":"ojs"},{"id":153,"value":"data_lifo = compute(st => st.pop(),\n st => st.pop(),\n false);","pinned":false,"hidden":false,"mode":"ojs"},{"id":192,"value":"md`### Recovered (using LIFO, recoveries applied before deaths)`","pinned":false,"hidden":false,"mode":"ojs"},{"id":155,"value":"chart(data_lifo.recArr)","pinned":false,"hidden":false,"mode":"ojs"},{"id":198,"value":"md`### Deceased (using LIFO, recoveries applied before deaths)`","pinned":false,"hidden":false,"mode":"ojs"},{"id":157,"value":"chart(data_lifo.decArr)","pinned":false,"hidden":false,"mode":"ojs"},{"id":146,"value":"md`## Setup and Code`","pinned":false,"hidden":false,"mode":"ojs"},{"id":458,"value":"computeIter = function(startData, numIter = iter_opts.n_iter) {\n let newData = startData;\n for (let i=0; i < numIter; i++) {\n newData = compute(\n (st, t) => removeWCDF(st, t, newData.cumDec),\n (st, t) => removeWCDF(st, t, newData.cumRec),\n true);\n };\n return newData;\n}","pinned":true,"hidden":false,"mode":"ojs"},{"id":109,"value":"$32","pinned":true,"hidden":false,"mode":"ojs"},{"id":266,"value":"removeWCDF = function(curArr, today, probCDF) {\n const maxDuration = Math.floor(today - curArr[0].day_in);\n \n let iter = 20;\n while (true) {\n // take a random active patient \n const idx = Math.floor(Math.random()*curArr.length);\n const dur = today - curArr[idx].day_in;\n // remove with probability decided by probCDF, linearly scaled using maxDuration to run faster\n if (--iter == 0 || decideExitWCDF(dur, probCDF, maxDuration)) {\n return curArr.splice(idx, 1)[0];\n }\n }\n }","pinned":true,"hidden":false,"mode":"ojs"},{"id":269,"value":"decideExitWCDF = function(duration, cumArr, maxDuration) {\n const maxProb = cumArr[Math.min(cumArr.length-1, maxDuration)];\n if (maxProb == 0) {\n return true;\n }\n const threshold = cumArr[Math.min(cumArr.length-1, duration)]/maxProb;\n return threshold <= 0 ? false : threshold >= 1 ? true : (Math.random() < threshold);\n }","pinned":true,"hidden":false,"mode":"ojs"},{"id":491,"value":"computeStats = function (arr, idx) {\n let sa = arr.sort((a,b) => a - b);\n return {\n days: idx,\n mean: d3.mean(sa),\n median: d3.quantileSorted(sa, 0.5),\n d025: d3.quantileSorted(sa, 0.025),\n d20: d3.quantileSorted(sa, 0.2),\n d80: d3.quantileSorted(sa, 0.8),\n d975: d3.quantileSorted(sa, 0.975)\n }\n}","pinned":true,"hidden":false,"mode":"ojs"},{"id":309,"value":"smoothen = function(data) {\n const ret = new Array(data.length);\n ret[0] = data[0];\n for (let i = 1; i < data.length-1; i++) {\n ret[i] = (data[i] + data[i-1] + data[i+1])/3.0;\n }\n ret[ret.length-1] = data[ret.length-1];\n return data;\n}","pinned":true,"hidden":false,"mode":"ojs"},{"id":230,"value":"dist2cum = function(distArr) {\n const cumArr = [];\n let sum = 0;\n for (let i=0; i