30:T5e0,scored = { let data = tweets // check if we want to include replies from radio button above .filter(d => usereplies == "no replies" ? d["Tweet text"][0] !== "@" : true) // filter tweets below a certain threshold .filter(d => d.impressions > impressionThreshold) // let's make a copy of the tweets for marking up with our scores .map(d => ({...d})); // first we sort by impressions and then give the top 5 tweets a score starting with 5 at the top data.sort((a,b) => b.impressions - a.impressions) data.forEach((d,i) => { if(i < 5) { d.impressionScore = 5 - i } else { d.impressionScore = 0 } }) // now we sort by engagement rate, but filter out any tweets with low impressions to get rid of noise // and then we give the top 5 tweets a score starting with 5 at the top data .sort((a,b) => b.engagementRate - a.engagementRate) .forEach((d, i) => { if(i < 5) { d.engagementScore = 5 - i } else { d.engagementScore = 0 } }) // finally we sort by user profile clicks, because we want to know what will drive follows data .sort((a,b) => b["user profile clicks"] - a["user profile clicks"]) .forEach((d, i) => { if(i < 5) { d.clickScore = 5 - i } else { d.clickScore = 0 } }) // total up the scores to get our top scoring tweets data.forEach(d => { d.totalScore = d.impressionScore + d.engagementScore + d.clickScore }) return data }31:T401,gridify( [ { element: htl.html`

Score: 5
Impressions: 24,320
Engagements: 517
Engagement Rate: 2.13%

`, cols: 2, styles: { "font-weight":"bold", "font-size":"16px", "font-family":"sans-serif", margin:"auto" }, }, { element: htl.html`

Score: 0
Impressions: 4,328
Engagements: 550
Engagement Rate: 21.71%

`, cols: 2, styles: { "font-weight":"bold", "font-size":"16px", "font-family":"sans-serif", margin:"auto" }, }, { element: frostyTweet, cols: 2, styles: { "max-width":"350px", margin:"auto", }, }, { element: yearTweet, cols: 2, styles: { "max-width":"350px", margin:"auto", "margin-top":"0px", }, }, ], // Full grid styles { columns: 4, styles: { padding: "10px", "max-width": isMobile ? "100%" : "700px" } } )32:T995,Plot.plot({ marks: [ Plot.link(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x1: "time", x2: "time", y1: flowerHeight === "engagementRate" ? 0 : 1, y2: flowerHeight, stroke: "green", opacity: 0.2 }), // profile clicks, blue Plot.dot(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: "time", y: flowerHeight, r: "profileClicks", fill: "#1DA1F2", stroke: "#FFFFFF", opacity: 0.75, dx: -10, // would be nice to make this dynamic based on radius dy: -10 }), // likes, red Plot.dot(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: "time", y: flowerHeight, r: "likes", fill: "red", stroke: "#FFFFFF", opacity: 0.75, dx: -10, dy: 10, }), // total score, yellow Plot.dot(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: "time", y: flowerHeight, r: d => d.mediaViews / 20, //could also use totalScore fill: "#FFCC00", stroke: "#FFFFFF", opacity: 0.75, dx: 10, dy: -10 }), // retweets, purple Plot.dot(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: "time", r: "retweets", y: flowerHeight, fill: "purple", stroke: "#FFFFFF", opacity: 0.75, dx: 10, dy: 10, }), // middle of the flower Plot.dot(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: "time", y: flowerHeight, r: 2, fill: "#FFFFFF", stroke: "#222222", opacity: 1, }), // tooltip Plot.dot(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: "time", // date y: flowerHeight, r: 5, stroke: "#FFFFFF", opacity: 0, title: (d) => `${d.tweetText.substring(0, 20)+ "..."} \n Impressions: ${d.impressions} \n Engagements: ${d.engagements} \n Engagement Rate ${d.engagementRate.toFixed(2) + "%"} \n Total Score: ${d.totalScore}`, }), ], height: isMobile ? 900 : 500, width: isMobile ? 400 : 1024, x: { domain: [earliestDate, latestDate], // axis: null, }, y: { type: flowerHeight === "engagementRate" ? "linear" : "log", label: flowerHeight === "engagementRate" ? `${flowerHeight}, in %` : flowerHeight, } })33:T953,Plot.plot({ marks: [ Plot.link(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x1: flowerHeight === "engagementRate" ? 0 : 1, x2: flowerHeight, y1: "time", y2: "time", stroke: "green", opacity: 0.2 }), // profile clicks, blue Plot.dot(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: flowerHeight, y: "time", r: "profileClicks", fill: "#1DA1F2", stroke: "#FFFFFF", opacity: 0.75, dx: -10, dy: -10 }), // likes, red Plot.dot(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: flowerHeight, y: "time", r: "likes", fill: "red", stroke: "#FFFFFF", opacity: 0.75, dx: -10, dy: 10, }), // total score, yellow Plot.dot(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: flowerHeight, y: "time", r: d => d.mediaViews / 20, // totalScore fill: "#FFCC00", // yellow stroke: "#FFFFFF", opacity: 0.75, dx: 10, dy: -10 }), // retweets, purple Plot.dot(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: flowerHeight, y: "time", r: "retweets", fill: "purple", stroke: "#FFFFFF", opacity: 0.75, dx: 10, dy: 10 }), // middle of the flower Plot.dot(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: flowerHeight, y: "time", r: 2, fill: "#FFFFFF", stroke: "#222222", opacity: 1, }), // tooltip Plot.dot(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: flowerHeight, // date y: "time", r: 5, stroke: "#FFFFFF", opacity: 0, title: (d) => `${d.tweetText.substring(0, 40)+ "..."} \n Likes: ${d.likes} \n Profile Clicks: ${d.profileClicks} \n Retweets ${d.retweets} \n Media Views: ${d.mediaViews}`, }), ], height: isMobile ? 800 : 900, width: isMobile ? 400 : 800, x: { type: flowerHeight === "engagementRate" ? "linear" : "log", label: flowerHeight === "engagementRate" ? `${flowerHeight}, in %` : flowerHeight, axis: "top" }, y: { domain: [earliestDate, latestDate], reverse: true, axis: "right" } })34:T6f5,Plot.plot({ marks: [ // likes Plot.vector(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: flowerHeight, y: "time", rotate: 0, length: "likes", // dx: 10, // dy: -10, // stroke: 3 }), // likes Plot.vector(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: flowerHeight, y: "time", rotate: 120, length: "profileClicks", // dx: 10, // dy: -10, // stroke: 3 }), // retweets Plot.vector(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: flowerHeight, y: "time", rotate: 240, length: "retweets", // dx: 10, // dy: 10 }), // // middle of the flower // Plot.dot(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { // x: flowerHeight, // y: "time", // r: 2, // fill: "#1DA1F2", // stroke: "#FFFFFF", // opacity: 1, // }), // tooltip Plot.dot(scoredWrangled.filter(d => d.impressions > chartImpressionThreshold), { x: flowerHeight, // date y: "time", r: 5, stroke: "#FFFFFF", opacity: 0, title: (d) => `${d.tweetText.substring(0, 40)+ "..."} \n Likes: ${d.likes} \n Profile Clicks: ${d.profileClicks} \n Retweets ${d.retweets} \n Media Views: ${d.mediaViews}`, }), ], height: isMobile ? 800 : 900, width: isMobile ? 400 : 800, background: "#222222", x: { type: flowerHeight === "engagementRate" ? "linear" : "log", label: flowerHeight === "engagementRate" ? `${flowerHeight}, in %` : flowerHeight, axis: "top" }, y: { domain: [earliestDate, latestDate], reverse: true, axis: "right" } })35:[{"id":"2f4abff5e02224286b419fcbf06c06dc463b2080b798ef607157ac64bc139410a3b6af80c5fe9efcd8fcae8fe29e691a7e56746c2c2ea6357844938396375459","href":"https://static.observableusercontent.com/files/2f4abff5e02224286b419fcbf06c06dc463b2080b798ef607157ac64bc139410a3b6af80c5fe9efcd8fcae8fe29e691a7e56746c2c2ea6357844938396375459","name":"tweet_activity_metrics_wiederkehra_20220101_20220201_en.csv","type":"text/csv","lastModified":1652877009817.482,"size":19971},{"id":"d0a67c8ff04f8647943be5f40b4a1d083b1a161b1e038306f65c3510e9286196a5dc08c691c7ad85f54d7b713539f70217060db2a3d528a0d152e8bf1448ac3f","href":"https://static.observableusercontent.com/files/d0a67c8ff04f8647943be5f40b4a1d083b1a161b1e038306f65c3510e9286196a5dc08c691c7ad85f54d7b713539f70217060db2a3d528a0d152e8bf1448ac3f","name":"tweet_activity_metrics_wiederkehra_20220201_20220301_en.csv","type":"text/csv","lastModified":1652898373771.754,"size":29038},{"id":"043ac335e4aa50e27b1b33b8b04858913d54693b169454a607aa12a2c90477ec0dd32e98ed68e861474a8b1089c8fa9edb2ef4855298e0dd4c99184ff6c1e4f8","href":"https://static.observableusercontent.com/files/043ac335e4aa50e27b1b33b8b04858913d54693b169454a607aa12a2c90477ec0dd32e98ed68e861474a8b1089c8fa9edb2ef4855298e0dd4c99184ff6c1e4f8","name":"tweet_activity_metrics_wiederkehra_20220401_20220501_en.csv","type":"text/csv","lastModified":1652898373736.781,"size":9326},{"id":"2f4abff5e02224286b419fcbf06c06dc463b2080b798ef607157ac64bc139410a3b6af80c5fe9efcd8fcae8fe29e691a7e56746c2c2ea6357844938396375459","href":"https://static.observableusercontent.com/files/2f4abff5e02224286b419fcbf06c06dc463b2080b798ef607157ac64bc139410a3b6af80c5fe9efcd8fcae8fe29e691a7e56746c2c2ea6357844938396375459","name":"tweet_activity_metrics_wiederkehra_20220101_20220201_en@1.csv","type":"text/csv","lastModified":1652898373750.678,"size":19971},{"id":"b141509c014d25f25f99266b4b43211151a0a17e7eb6527e4df6524f5e74db185eb15608ae1dbef8e7051dbe1570e2cdd93908866807e6aa19120cae21073013","href":"https://static.observableusercontent.com/files/b141509c014d25f25f99266b4b43211151a0a17e7eb6527e4df6524f5e74db185eb15608ae1dbef8e7051dbe1570e2cdd93908866807e6aa19120cae21073013","name":"tweet_activity_metrics_wiederkehra_20220301_20220401_en.csv","type":"text/csv","lastModified":1652898373790.213,"size":37103},{"id":"aff7d93ba2116db44ee0a36712d911729b1c48b8f4f579596bf88c23d22860faf09d40bf9c8dbc6f996db335b7a22681e13f3b93e14bf2412ec3b5f4084d15cd","href":"https://static.observableusercontent.com/files/aff7d93ba2116db44ee0a36712d911729b1c48b8f4f579596bf88c23d22860faf09d40bf9c8dbc6f996db335b7a22681e13f3b93e14bf2412ec3b5f4084d15cd","name":"tweet_activity_metrics_wiederkehra_20220501_20220519_en.csv","type":"text/csv","lastModified":1652898935407.108,"size":2630}] 2e:{"title":"Tweet content analysis","cells":[{"id":0,"value":"# Tweet content analysis\nForked from [Ian Johnson's notebook](https://observablehq.com/@observablehq/tweet-content-niche-analysis) translating [Tobias Zwingmann's Twitter thread](https://twitter.com/ztobi/status/1490414903269670917?ref_src=twsrc%5Etfw%7Ctwcamp%5Etweetembed%7Ctwterm%5E1490414903269670917%7Ctwgr%5E%7Ctwcon%5Es1_&ref_url=https%3A%2F%2Fwiederkehra.static.observableusercontent.com%2Fnext%2Fworker-fc08426a.html). ","pinned":false,"hidden":false,"mode":"md"},{"id":150,"value":"---\n","pinned":false,"hidden":false,"mode":"md"},{"id":232,"value":"### Import data from Twitter analytics","pinned":false,"hidden":false,"mode":"md"},{"id":234,"value":"I used data from May 19, 2022 through the beginning of the year.","pinned":false,"hidden":false,"mode":"md"},{"id":564,"value":"jan = FileAttachment(\"tweet_activity_metrics_wiederkehra_20220101_20220201_en@1.csv\").csv( {typed: true} ) ","pinned":false,"hidden":false,"mode":"ojs"},{"id":566,"value":"feb = FileAttachment(\"tweet_activity_metrics_wiederkehra_20220201_20220301_en.csv\").csv({typed: true})","pinned":false,"hidden":false,"mode":"ojs"},{"id":568,"value":"mar = FileAttachment(\"tweet_activity_metrics_wiederkehra_20220301_20220401_en.csv\").csv({typed: true})","pinned":false,"hidden":false,"mode":"ojs"},{"id":570,"value":"apr = FileAttachment(\"tweet_activity_metrics_wiederkehra_20220401_20220501_en.csv\").csv({typed: true})","pinned":false,"hidden":false,"mode":"ojs"},{"id":819,"value":"may = FileAttachment(\"tweet_activity_metrics_wiederkehra_20220501_20220519_en.csv\").csv({typed: true})","pinned":false,"hidden":false,"mode":"ojs"},{"id":814,"value":"Merge 'em all together","pinned":false,"hidden":false,"mode":"md"},{"id":573,"value":"tweetData = [...jan, ...feb, ...mar, ...apr, ...may];","pinned":false,"hidden":false,"mode":"ojs"},{"id":126,"value":"viewof usereplies = Inputs.radio([\"include replies\", \"no replies\"], {label: \"\", value: \"include replies\"})","pinned":false,"hidden":false,"mode":"ojs"},{"id":45,"value":"impressionThreshold = 10 // change this number to filter out ","pinned":false,"hidden":false,"mode":"ojs"},{"id":67,"value":"Inputs.table(scored, {\n columns: [\n \"totalScore\",\n \"Tweet text\",\n \"Tweet permalink\",\n \"impressions\",\n \"engagements\",\n \"engagement rate\",\n \"retweets\",\n \"replies\",\n \"likes\",\n \"user profile clicks\",\n \"url clicks\",\n \"time\",\n \"impressionScore\",\n \"engagementScore\",\n \"clickScore\",\n \"Tweet id\",\n ],\n width: {\n totalScore: 20,\n \"Tweet text\": 100,\n \"Tweet permalink\": 55,\n impressions: 65,\n engagements: 65,\n \"engagement rate\": 65,\n retweets: 45,\n replies: 45,\n likes: 45,\n \"user profile clicks\": 45,\n \"url clicks\": 45,\n time: 140\n },\n format: {\n \"Tweet permalink\": (d) => htl.html`${d}`,\n \"Tweet text\": d => d.length > 50 ? d.slice(0,50) + \"...\" : d,\n \"engagement rate\": d => d3.format(\"0.2%\")(d)\n },\n sort: \"totalScore\", reverse: true \n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":309,"value":"### Some definitions\nEngagement is the total number of times a user interacted with your tweet – including retweets, replies, follows, likes, links, cards, hashtags, embedded media, username, profile photo, or tweet expansion.\n
Impressions are the times a user is served your tweet in timeline or search results.\nEngagement Rate is equal to your tweets' engagement divided by the number of impressions.\n([source](https://help.twitter.com/en/managing-your-account/using-the-tweet-activity-dashboard))","pinned":false,"hidden":false,"mode":"md"},{"id":321,"value":"---","pinned":false,"hidden":false,"mode":"md"},{"id":159,"value":"### Scored tweets","pinned":false,"hidden":false,"mode":"md"},{"id":971,"value":"Scoring is translated by Ian, from Tobias's post","pinned":false,"hidden":false,"mode":"md"},{"id":20,"value":"$30","pinned":false,"hidden":false,"mode":"ojs"},{"id":306,"value":"A little wrangling for my own sanity (renaming and pulling only what we want)","pinned":false,"hidden":false,"mode":"md"},{"id":268,"value":"scoredWrangled = scored.map(el => ({\n // fix these spaced names\n tweetID: el[\"Tweet id\"], \n tweetPerma: el[\"Tweet permalink\"], \n tweetText: el[\"Tweet text\"], \n engagementRate: parseFloat(el[\"engagement rate\"]) * 100, // make percentage\n impressions: parseFloat(el.impressions), // value not string\n engagements: parseFloat(el.engagements), // value not string\n totalScore: el.totalScore,\n time: new Date(el[\"time\"]),\n likes: el[\"likes\"],\n retweets: el[\"retweets\"],\n profileClicks: el[\"user profile clicks\"],\n mediaViews: el[\"media views\"],\n}));","pinned":false,"hidden":false,"mode":"ojs"},{"id":907,"value":"Without spending much – or any – time looking at the way the scoring works, let's just see if we can find out through seeing how my tweets perform via Twitter's statistics (impressions, engagements, engagement rates) and the scoring through visualization.","pinned":false,"hidden":false,"mode":"md"},{"id":389,"value":"Sized by engagement","pinned":false,"hidden":false,"mode":"md"},{"id":346,"value":"Plot.plot({\n marks: [\n Plot.dot(scoredWrangled, {\n x: \"impressions\", // served your tweet\n r: \"engagements\", // interacted with your tweet\n fill: \"#1DA1F2\",\n title: (d) => \n `${d.tweetText.substring(0, 20)+ \"...\"} \\n Impressions: ${d.impressions} \\n Engagements: ${d.engagements} \\n Engagement Rate ${d.engagementRate.toFixed(2) + \"%\"} \\n Total Score: ${d.totalScore}`,\n stroke: \"#FFFFFF\",\n opacity: 0.6,\n })\n ],\n height: isMobile ? 900 : 200,\n width: isMobile ? 400 : 1024,\n x: { \n // tickFormat: d => d.substring(0, 2)\n // axis: null,\n // reverse: true,\n label: \"IMPRESSIONS (served)\"\n }\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":525,"value":"Sized by impressions","pinned":false,"hidden":false,"mode":"md"},{"id":522,"value":"Plot.plot({\n marks: [\n Plot.dot(scoredWrangled, {\n x: \"engagements\", // served your tweet\n r: \"impressions\", // interacted with your tweet\n fill: \"#1DA1F2\",\n title: (d) => \n `${d.tweetText.substring(0, 20)+ \"...\"} \\n Impressions: ${d.impressions} \\n Engagements: ${d.engagements} \\n Engagement Rate ${d.engagementRate.toFixed(2) + \"%\"} \\n Total Score: ${d.totalScore}`,\n stroke: \"#FFFFFF\",\n opacity: 0.6,\n })\n ],\n height: isMobile ? 900 : 200,\n width: isMobile ? 400 : 1024,\n x: { \n // tickFormat: d => d.substring(0, 2)\n // axis: null,\n // reverse: true,\n label: \"ENGAGEMENTS (interaction)\"\n }\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":919,"value":"The two charts above look at engagement and impressions, flipping the sizing and axis between each other.\n- The same tweet ranks highest in both charts.\n- The first chart, arranged by impressions, has a _litte more_ distribution than the second chart. Tweets that were served to more people seem to get more engagement (smaller circles for lower impressions).\n- However, the second chart, arranged by engagements, shows better that even though some tweets were served a lot more (yielding higher impressions), it doesn't mean they got a lot of engagement. There are a cluster of larger circles around and below 500 engagements that, shown by their size, had a lot of impressions. ","pinned":false,"hidden":false,"mode":"md"},{"id":928,"value":"Now let's look at engagement rate (engagements / impressions). Let's say I had a tweet that was promoted a lot in people's timeline and search, but not many people interacted with it. That tweet would find itself more toward the left, in the lower numbers of the chart above. Conversely, a tweet that most of the people who saw it, interacted with it, would land more toward the upper end of the chart. ","pinned":false,"hidden":false,"mode":"md"},{"id":535,"value":"viewof metric = Inputs.select([\"engagements\", \"impressions\", \"totalScore\"], {label: \"Size chart below by\"})","pinned":false,"hidden":false,"mode":"ojs"},{"id":529,"value":"Plot.plot({\n marks: [\n Plot.dot(scoredWrangled, {\n x: d => d.engagementRate / 100, // dividing by 100 so I can format below\n r: metric, // chosen in the input above the chart\n fill: \"#1DA1F2\",\n title: (d) => \n `${d.tweetText.substring(0, 20)+ \"...\"} \\n Impressions: ${d.impressions} \\n Engagements: ${d.engagements} \\n Total Score: ${d.totalScore} \\n Engagement Rate: ${d.engagementRate.toFixed(2)}%`,\n stroke: \"#FFFFFF\",\n opacity: 0.6,\n })\n ],\n height: isMobile ? 900 : 200,\n width: isMobile ? 400 : 1024,\n x: { \n tickFormat: d3.format(\",.1%\"),\n // axis: null,\n grid: true,\n label: \"ENGAGEMENT RATE (engagements / impressions)\"\n },\n // y: {\n // grid: true,\n // }\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":558,"value":"---","pinned":false,"hidden":false,"mode":"md"},{"id":933,"value":"Now let's size the tweets by our score. Below, the blue dots rendered in the chart are dots where a tweet has been assigned a score greater than 0, the gray dots have a score of 0. In my case, only seven of my 122 tweets since the beginning of the year have a score greater than 0.","pinned":false,"hidden":false,"mode":"md"},{"id":850,"value":"viewof yAxis = Inputs.select([\"engagements\", \"engagementRate\", \"totalScore\", \"profileClicks\"], {label: \"Metric on the y axis\"})","pinned":false,"hidden":false,"mode":"ojs"},{"id":391,"value":"Sized by total score","pinned":false,"hidden":false,"mode":"md"},{"id":236,"value":"Plot.plot({\n marks: [\n Plot.dot(scoredWrangled.filter(d => d.totalScore === 0), {\n x: \"impressions\",\n y: yAxis,\n fill: \"#cccccc\",\n stroke: \"#ffffff\", \n // r: 1,\n opacity: 0.8,\n }),\n Plot.dot(scoredWrangled, {\n x: \"impressions\",\n y: yAxis,\n fill: \"#1DA1F2\", // Twitter blue\n stroke: \"#FFFFFF\", \n r: \"totalScore\",\n opacity: 1,\n title: (d) => \n `${d.tweetText.substring(0, 20)+ \"...\"} \\n Impressions: ${d.impressions} \\n Engagements: ${d.engagements} \\n Engagement Rate ${d.engagementRate.toFixed(2) + \"%\"} \\n Total Score: ${d.totalScore}`,\n }),\n ],\n y: { \n label: yAxis === \"engagementRate\" ? yAxis.toUpperCase() + \" (%)\" : yAxis.toUpperCase(),\n },\n x: { \n label: \"IMPRESSIONS\"\n }\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":964,"value":"- If we show engagement rate on the y axis, we can see almost immediately that it isn't rated very high in our scoring methodology. The tweets with the highest engagement rates don't even rank.\n- Putting engagements on the y shows us how this stat seems to pull the largest weight in our scoring. The tweet with the highest engagement also has the highest score. But we also see there is a little nuance among the other scored tweets (some have lower engagements but have higher scores among each other).\n- With profile clicks on the y, we get an answer to why the tweet with the lowest TKTK ","pinned":false,"hidden":false,"mode":"md"},{"id":223,"value":"---","pinned":false,"hidden":false,"mode":"md"},{"id":505,"value":"This year so far, my tweets with the highest impressions and engagements was about [a project I did about the NFL while at FiveThirtyEight](https://projects.fivethirtyeight.com/tom-brady-touchdown-passes/). The tweet with the highest engagement rate was, as expected, the one that linked to the project (below, right). The tweet with the most impressions and engagement (below, left) is the first tweet in a thread about the process of the project. ","pinned":false,"hidden":false,"mode":"md"},{"id":414,"value":"gridify(\n [\n {\n element: htl.html`
Highest score (14)
Most impressions (66,086)
Most engagements (5,640)
`,\n cols: 2,\n styles: { \n \"font-weight\":\"bold\",\n \"font-size\":\"16px\",\n \"font-family\":\"sans-serif\",\n margin:\"auto\"\n },\n },\n {\n element: htl.html`
Low score (0)
Highest engagement rate (15.36%)
`,\n cols: 2,\n styles: { \n \"font-weight\":\"bold\",\n \"font-size\":\"16px\",\n \"font-family\":\"sans-serif\",\n margin:\"auto\",\n \"margin-top\":\"0px\"\n },\n },\n {\n element: mostImpressed,\n cols: 2,\n styles: { \n margin:\"auto\",\n \"margin-top\":\"0px\"\n },\n },\n {\n element: mostEngaged,\n cols: 2,\n styles: { \n margin:\"auto\",\n },\n },\n ],\n // Full grid styles\n {\n columns: 4,\n styles: {\n padding: \"10px\",\n \"max-width\": isMobile ? \"100%\" : \"700px\"\n }\n }\n)","pinned":false,"hidden":false,"mode":"ojs"},{"id":473,"value":"Two other tweets help me understand a little more about how these stats can tell different stories. Below are tweets where there are photos, specifically, photos of me. \n\nThe on the left is the 5th most impressed tweet – meaning it was served to lots of people via their timeline or search – but the engagement rate is fairly low. We end up with a score of 5 (out of my max of 14) for this tweet. \n\nOn the right, a tweet with several photos has the third highest engagement rate of all my tweets since the beginning of the year. But, it has 18% of the impressions of the tweet on the left and a resulting score of 0. ","pinned":false,"hidden":false,"mode":"md"},{"id":482,"value":"$31","pinned":false,"hidden":false,"mode":"ojs"},{"id":905,"value":"We can conclude that the scoring weights tweets with higher impressions more heavily than tweets with higher engagement rates. ","pinned":false,"hidden":false,"mode":"md"},{"id":400,"value":"---","pinned":false,"hidden":false,"mode":"md"},{"id":179,"value":"### Unfiltered and unscored tweets","pinned":false,"hidden":false,"mode":"md"},{"id":78,"value":"tweets = tweetData","pinned":false,"hidden":false,"mode":"ojs"},{"id":25,"value":"Inputs.table(tweets, {\n columns: [\n \"impressions\", \"engagements\", \"engagement rate\", \"retweets\", \"replies\", \"likes\", \"user profile clicks\", \"url clicks\", \"time\", \"Tweet permalink\", \"Tweet text\", \"Tweet id\"\n ],\n width: {\n \"impressions\": 65,\n \"engagements\": 65,\n \"engagement rate\": 65,\n \"retweets\": 45,\n \"replies\": 45,\n \"likes\": 45,\n \"user profile clicks\": 45,\n \"url clicks\": 45,\n \"Tweet permalink\": 55,\n \"time\": 140,\n },\n format: {\n \"Tweet permalink\": d => htl.html`${d}`,\n \"Tweet text\": d => d.length > 50 ? d.slice(0,50) + \"...\" : d,\n \"engagement rate\": d => d3.format(\"0.2%\")(d)\n },\n sort: \"impressions\", reverse: true\n})","pinned":false,"hidden":false,"mode":"ojs"},{"id":935,"value":"nonZero = scoredWrangled.filter(d => d.totalScore > 0)","pinned":false,"hidden":false,"mode":"ojs"},{"id":344,"value":"### Imports","pinned":false,"hidden":false,"mode":"md"},{"id":252,"value":"import {Plot} from \"@mkfreeman/plot-tooltip\"","pinned":false,"hidden":false,"mode":"ojs"},{"id":350,"value":"import {isMobile} from \"@wiederkehra/how-fast-to-podium\"","pinned":false,"hidden":false,"mode":"ojs"},{"id":397,"value":"import {tweet} from \"@mbostock/tweet\"","pinned":false,"hidden":false,"mode":"ojs"},{"id":410,"value":"import {gridify} from \"@observablehq/gridify\"","pinned":false,"hidden":false,"mode":"ojs"},{"id":1110,"value":"import {textcolor} from \"@observablehq/text-color-annotations-in-markdown\"","pinned":false,"hidden":false,"mode":"ojs"},{"id":562,"value":"---","pinned":false,"hidden":false,"mode":"md"},{"id":442,"value":"### Variables","pinned":false,"hidden":false,"mode":"md"},{"id":478,"value":"yearTweet = tweet(\"1477767022406635521\")","pinned":true,"hidden":false,"mode":"ojs"},{"id":476,"value":"frostyTweet = tweet(\"1486726637945176064\")","pinned":true,"hidden":false,"mode":"ojs"},{"id":406,"value":"mostImpressed = tweet(\"1480928238465601542\")","pinned":true,"hidden":false,"mode":"ojs"},{"id":402,"value":"mostEngaged = tweet(\"1480928308028182532\")","pinned":true,"hidden":false,"mode":"ojs"},{"id":983,"value":"### Chart explorations","pinned":false,"hidden":false,"mode":"md"},{"id":994,"value":"viewof flowerHeight = Inputs.select([\"impressions\", \"engagements\", \"engagementRate\", \"profileClicks\"], {label: \"Height of flowers should be\"})","pinned":false,"hidden":false,"mode":"ojs"},{"id":1097,"value":"chartImpressionThreshold = 0","pinned":false,"hidden":false,"mode":"ojs"},{"id":1116,"value":"$${textcolor(\"likes\", \"red\")} ${textcolor(\"profile clicks\", \"#1DA1F2\")} ${textcolor(\"retweets\", \"purple\")} ${textcolor(\"media views\", \"#FFCC00\")}","pinned":false,"hidden":false,"mode":"md"},{"id":985,"value":"$32","pinned":false,"hidden":false,"mode":"ojs"},{"id":1146,"value":"$33","pinned":false,"hidden":false,"mode":"ojs"},{"id":1250,"value":"$34","pinned":false,"hidden":false,"mode":"ojs"},{"id":1063,"value":"latestDate = d3.max(scoredWrangled, d => d.time)","pinned":false,"hidden":false,"mode":"ojs"},{"id":1068,"value":"earliestDate = new Date(d3.min(scoredWrangled, d => d.time) - 500000000)","pinned":false,"hidden":false,"mode":"ojs"}],"files":"$W35","stdlib":"1","titling":"auto"}