2e:null 2f:["$","title",null,{"children":"Covid-19: China vs other countries | J Kae | Observable"}] 34:T742,function aggregate_by_country (hurdle) { //hurdle is the minimum of confirmed cases to start counting var data_input = denormalized(); var new_array = []; var i=0; while (i ( e.country === data_input[i].country && e.date_formatted === data_input[i].date_formatted )); //if target_update_index = -1 then add a new row with the information if (target_update_index == -1) { new_array.push( { "country": data_input[i].country, "date_formatted": data_input[i].date_formatted, "date": data_input[i].date, "confirmed": data_input[i].confirmed, "day" : 0, } ) } else { //add the confirmed cases to the matching object with some errorchecking if (isNaN(data_input[i].confirmed)) {} else { new_array[target_update_index].confirmed = Number(data_input[i].confirmed) + Number( new_array[target_update_index].confirmed); } } i++ } //Now, let's add the days since confirmed exceeded 0 i = 0; while (i < new_array.length) { if (new_array[i].confirmed > hurdle) //only if confirmed cases exceeded 0 { if (i==0) { new_array[i].day = 1 } // different if the first element else { if (new_array[i].country == new_array[i-1].country) ///needs to be day=1 if the new row is a new country { new_array[i].day = 1 + Number(new_array[i-1].day); // otherwise, just increment the day } else { new_array[i].day = 1; } } } i++; } return new_array; } 35:T496,vl.markLine() .data(df) .mark("line") .transform( [ {"sortby": [ "country", "day" ] }, {"window": [ { "op": "lead", "field": "confirmed", "as": "confirmed_tomorrow" } ] }, {"calculate": "datum.confirmed_tomorrow - datum.confirmed", "as": "change"}, {"filter": {"and": [ "datum.day <= 21 && datum.day > 0", "datum.change > 0", //the last day seems to have a major drop-off which doesn't make sense {"or" : [ {"field": "country", "equal": "Mainland China"}, {"field": "country", "equal": "South Korea"}, {"field": "country", "equal": "Italy"}, {"field": "country", "equal": "Iran"} ] } ] } } ]) .encode({ "x": {"field": "day", "type": "nominal"}, "y": { "field": "change", "type": "quantitative" }, "color": { "field": "country", "type": "nominal" }, } ) .render() 36:T492,vl.markLine() .data(df) .mark("line") .transform( [ {"sortby": [ "country", "day" ] }, {"window": [ { "op": "lead", "field": "confirmed", "as": "confirmed_tomorrow" } ] }, {"calculate": "(datum.confirmed_tomorrow/datum.confirmed-1)*100", "as": "percent_change"}, {"filter": {"and": [ "datum.day <= 21 && datum.day > 0", "datum.percent_change > 1",//"isValid(datum.confirmed_tomorrow)", {"or" : [ {"field": "country", "equal": "Mainland China"}, {"field": "country", "equal": "South Korea"}, {"field": "country", "equal": "Italy"}, {"field": "country", "equal": "Iran"} ] } ] } } ]) .encode({ "x": {"field": "day", "type": "nominal"}, "y": { "field": "percent_change", "type": "quantitative" }, "color": { "field": "country", "type": "nominal" }, } ) .render() 37:T4e6,vl.markLine() .data(df) .mark("line") .transform( [ {"sortby": [ "country", "day" ] }, {"window": [ { "op": "lead", "field": "confirmed", "as": "confirmed_tomorrow" } ] }, {"calculate": "(datum.confirmed_tomorrow / datum.confirmed -1 )*100", "as": "percent_change"}, {"filter": {"and": [ "datum.day <= 16 && datum.day > 0", "datum.percent_change > 0", "datum.confirmed < 10000",//"isValid(datum.confirmed_tomorrow)", {"or" : [ {"field": "country", "equal": "Mainland China"}, {"field": "country", "equal": "South Korea"}, {"field": "country", "equal": "Italy"}, {"field": "country", "equal": "Iran"} ] } ] } } ]) .encode({ "x": { "field": "confirmed", "type": "quantitative" }, "y": { "field": "percent_change", "type": "quantitative" }, "color": { "field": "country", "type": "nominal" }, } ) .render() 38:T530,function denormalized () { var data = []; //placeholder for denormalized data to be returned var i = 0; // counter for location //counters to iterate days var startDate = new Date(2020, 0, 22); var timespan = Object.keys(covid_data.confirmed.locations[0].history).length; // number of days in the dataset // iterate across locations while (i < covid_data.confirmed.locations.length) { var j = 0; while (j 50 && datum.day <= 25\"\n })\n .encode({\n \"x\": {\"field\": \"day\", \"type\": \"nominal\"},\n \"y\": {\n \"field\": \"confirmed\",\n \"type\": \"quantitative\"\n },\n \"color\": {\"field\": \"country\", \"type\": \"nominal\"},\n}\n \n )\n .render()","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":877,"value":"md`To increase readability further, I am plotting it on a log scale`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":878,"value":"vl.markLine()\n .data(df)\n .mark(\"point\")\n .transform({\n \"filter\": \"datum.confirmed > 50 && datum.day <= 25\"\n })\n .encode({\n \"x\": {\"field\": \"day\", \"type\": \"nominal\"},\n \"y\": {\n \"field\": \"confirmed\",\n \"type\": \"quantitative\",\n \"scale\": {\"type\": \"log\"}\n },\n \"color\": {\"field\": \"country\", \"type\": \"nominal\"},\n}\n \n )\n .render()","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":830,"value":"vl.markLine()\n .data(df)\n .mark(\"point\")\n .transform({\n \"filter\": \n {\"and\": [\"datum.day <= 21 && datum.day > 0\"\n ,\n {\"or\" : \n [\n {\"field\": \"country\", \"equal\": \"Mainland China\"},\n {\"field\": \"country\", \"equal\": \"South Korea\"},\n {\"field\": \"country\", \"equal\": \"Italy\"},\n {\"field\": \"country\", \"equal\": \"Iran\"}\n ]}\n ]\n }})\n .encode({\n \"x\": {\"field\": \"day\", \"type\": \"nominal\"},\n \"y\": {\n \"field\": \"confirmed\",\n \"type\": \"quantitative\",\n \"scale\": {\"type\": \"log\"}\n },\n \"color\": {\"field\": \"country\", \"type\": \"nominal\"},\n}\n \n )\n .render()","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":827,"value":"md`So far, the growth rates look somewhat comparable (and scarily exponential in the early stages). However, on later days the growth rate clearly seems to flatten. Let's focus on the three countries with the most confirmed cases: Mainland China, South Korea, Iran, and Italy next. And I'll cut it at 3 weeks (since only China exceeded my hurdle rate for longer than 3 weeks).`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":901,"value":"md`Apart from the different starting points (China's first report counts ~500 confirmed cases), the rates look very similar, with South Korea on a slightly flatter trajectory. Now, let's look at the absolute change over this timeframe`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":903,"value":"$35","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":958,"value":"md`Obviously, China is far highest on absolute change over the first 16 days, stabilizing new infections after about two weeks and then significantly reducing them. Next, let's look at relative growth rates.`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":959,"value":"$36","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1014,"value":"md`This makes it even clearer how China significantly reduced growth rates over time. Luckily, other countries have achieved the same (and probably not just because the virus ran out of \"hosts\", as the population in these countries is much bigger than the current count of COVID19 cases).`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1016,"value":"md`Finally, I am trying to adjust for the fact that China started \"counting late\" (i.e., Day 1 was at about 500 cases). Let me plot % change vs # of confirmed cases.`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":987,"value":"$37","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":1018,"value":"md`What we can see here is that early growth is obviously pretty high. However, starting at around ~2,000 cases, the growth rates of all countries plotted are dropping below China. Obviously, these countries had \"more time\" to prepare for what was coming, and could look to China for how to tackle it. However, I am surprised that these countries have managed to come in below China.`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":130,"value":"md`\n## Appendix`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":215,"value":"import {vl} from '@vega/vega-lite-api'","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":218,"value":"import {printTable} from '@uwdata/data-utilities'","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2,"value":"url=\"https://coronavirus-tracker-api.herokuapp.com/all\"\n\n","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":117,"value":"covid_data= (await fetch(url)).json()","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":584,"value":"md`My helper function to transpose the json from the API to a denormalized table as shown above`","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":395,"value":"$38","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"}],"files":"$W39","stdlib":"1","titling":"auto"}} 33:"$32:body"