31:T1724,display = html` ${(ageAdjust ? adjusted : favourites) .slice(0, numberOfResults) .map((element, index) => { return ``; }) .join("")}

${index + 1}.		${element.text}
		${Math.round( Math.floor(element.count) )} faves \| comments

`32:[{"id":"c3788f2414caf2e607c2116b07828101f005928d4e08bb20f83b9813e05706267c2332948a98138d2467a38e9cb01f7fb9096939f0b7808c651d383361867a8d","href":"https://static.observableusercontent.com/files/c3788f2414caf2e607c2116b07828101f005928d4e08bb20f83b9813e05706267c2332948a98138d2467a38e9cb01f7fb9096939f0b7808c651d383361867a8d","name":"hn3.csv.gz","type":"application/gzip","lastModified":1734873343627.131,"size":1986912}] 2f:{"title":"Most favorited Hacker News Posts","cells":[{"id":50,"value":"md`# Most favorited Hacker News Posts\n## 2020-09-01 to 2024-12-21\n\nI need some good technical reading recs for the vacation. Time for an update of the *[Most favorited Hacker News Posts of All Time](https://observablehq.com/@tomlarkworthy/hacker-favourites-analysis)*. If many hackers take the time to bookmark the same article, it is a strong indicator that it is worth reading.\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":409,"value":"viewof ageAdjust = checkbox({\n options: [{ value: \"toggle\", label: \"Adjust for recency bias?\" }],\n value: \"toggle\"\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":247,"value":"viewof includeInternals = checkbox({\n options: [{ value: \"toggle\", label: \"Include Tell/Ask HN?\" }],\n value: false\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":58,"value":"$31","pinned":false,"hidden":false,"mode":"ojs"},{"id":538,"value":"viewof numberOfResults = slider({\n min: 10,\n max: 1000,\n step: 10,\n value: 50,\n title: \"Number of results\"\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":531,"value":"The dataset is biased towards more recent favourites, but you can adjust for that.","pinned":false,"hidden":false,"mode":"md"},{"id":439,"value":"Plot.plot({\n title: \"Quantity of favourites over time\",\n marks: [\n Plot.ruleY([0]),\n Plot.lineY(favourites, Plot.binX({ y: \"count\" }, { x: \"time\", tip: true })),\n Plot.lineY(\n d3\n .range(new Date(\"2020-09-01\"), new Date(\"2024-12-01\"), 1e8)\n .map((d) => ({ x: new Date(d), y: regression.predict(d) })),\n { x: \"x\", y: \"y\", stroke: \"red\", strokeDasharray: [4, 4] }\n )\n ]\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":21,"value":"favourites = d3\n .csvParse(await unzip(await FileAttachment(\"hn3.csv.gz\")))\n .filter((value) => !value.link.startsWith(\"/next/item\") || includeInternals)\n .map((d) => ({\n ...d,\n count: d.count,\n time: new Date(d.time)\n }))","pinned":false,"hidden":false,"mode":"ojs"},{"id":450,"value":"bins = {\n const channels = Plot.lineY(\n favourites,\n Plot.binX({ y: \"count\" }, { x: \"time\", tip: true })\n ).initialize().channels;\n\n return channels.x.value.map((x, i) => ({\n time: x,\n count: channels.y.value[i]\n }));\n}","pinned":false,"hidden":false,"mode":"ojs"},{"id":426,"value":"regression = reg\n .regressionPoly()\n .x((d) => d.time)\n .y((d) => d.count)\n .order(2)(bins)","pinned":false,"hidden":false,"mode":"ojs"},{"id":520,"value":"adjusted = favourites\n .map((f) => ({\n ...f,\n count: (f.count / regression.predict(f.time)) * 1000\n }))\n .sort((a, b) => b.count - a.count)","pinned":false,"hidden":false,"mode":"ojs"},{"id":583,"value":"### Methodology\n\n- Pick top 10k user accounts by comment count\n- Fetch the first page of their favourites.\n- Count how often a link was favourited (ignore links older than 2020-09-01)\n- To adjust for recency bias we fit a curve to the score totals in bucketed time ranges, then use that curve to divide each link by its recency boost. ","pinned":false,"hidden":false,"mode":"md"},{"id":380,"value":"favourites","pinned":false,"hidden":false,"mode":"ojs"},{"id":278,"value":"import { checkbox, slider } from \"@jashkenas/inputs\"","pinned":false,"hidden":false,"mode":"ojs"},{"id":295,"value":"import { exporter } from \"@tomlarkworthy/exporter\"","pinned":false,"hidden":false,"mode":"ojs"},{"id":298,"value":"exporter()","pinned":false,"hidden":false,"mode":"ojs"},{"id":354,"value":"unzip = async (attachment) =>\n await new Response(\n (await attachment.stream()).pipeThrough(new DecompressionStream(\"gzip\"))\n ).text()","pinned":false,"hidden":false,"mode":"ojs"},{"id":429,"value":"reg = require(\"d3-regression@1\")","pinned":false,"hidden":false,"mode":"ojs"}],"files":"$W32","stdlib":"1","titling":"auto"}