30:T460,html`

Literate Epidemiology Reporting with Governance

Jonas S Almeida, Montse, ...

Abstract

In this example of Box based data management to generate live reports for the Confluence project. The critical test if if we can assemble data from the Box source in real time and without any (absolutely none!) breach of Box's data access governance. In this specific report we'll load Age data from different files and aggregate them in a single report. If successful, it would establish the feasability of complex realt-time reports in a popular literate programming platform. Observable Notebooks are the Web equivalent of "Jupyter Notebooks", with all the advantages of the Web - the data stays within the governance scope of the user and no downloads or installations are required. An interresting side effect of these advantages is that these reports can be assembled in any platform, including mobile devices wll the way to smart phones.

`31:T411,ageDistribution=async function(auth,div){ div=div||document.getElementById('computationDiv') // notice target div is the second variable, with a default value div.textContent='retrieving raw files ...' let getFile=async function(id){ return (await fetch(`https://api.box.com/2.0/files/${id}/content`,{ method:'GET', headers:{ Authorization:"Bearer "+auth.access_token } })).text() } function txt2json(txt){ let arr = txt.split(/\n/).map(r=>r.split(/\s/)) let aa = arr[0] // fieldnames / attributes let dt ={} // data aa.forEach(a=>{dt[a]=[]}) // arr.slice(1).forEach((vv,i)=>{ aa.forEach((a,j)=>{dt[a][i]=vv[j]}) }) //data.dt[k]=dt return dt } let cores={ USRT:'418080151298', PLCO:'418097175569', //PBCS:'418087762189', //AHS:'418088787178' } Object.keys(cores).forEach(k=>{ console.log('reading '+k) getFile(cores[k]) .then(txt2json) .then(dt=>{ data.dt[k]=dt plotIt(dt,k) }) }) }32:[] 2e:{"title":"ConfluenceAge","cells":[{"id":0,"value":"md`# ConfluenceAge`","pinned":false,"hidden":false,"mode":"ojs"},{"id":4,"value":"$30","pinned":false,"hidden":false,"mode":"ojs"},{"id":33,"value":"html`\n

Methods

REST APIs
OAUTH2
CSS

\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":38,"value":"html`\n

CSS

Document styling can be defined within the document or imported from an external source, for example, to comply with a publisher's normative preferences. The style for this specific report can be inspected by clicking on the \"^...\" that appears to the left of this paragraph when it is selected or when the mouse rolls over it.

\n\n`","pinned":false,"hidden":false,"mode":"ojs"},{"id":226,"value":"html`\n

Computation

The computation of the age distributions requires an authenticated bearar token to have access to the private data. That function is defined here and can be imported by itself in other Observable Notebooks, where it will produce the same plot for the data accessible with the bearer token of the remote user.

\n`","pinned":true,"hidden":false,"mode":"ojs"},{"id":253,"value":"html`\n

Data

If we do it right, the global data variable will be populated here after authentication but it won't be available for importing elsewhere.

\n`","pinned":true,"hidden":false,"mode":"ojs"},{"id":256,"value":"data={return {dt:{}}}","pinned":true,"hidden":false,"mode":"ojs"},{"id":272,"value":"html`\n

Ploting age distribution

plotly will be used...

\n`","pinned":true,"hidden":false,"mode":"ojs"},{"id":278,"value":"plotly=require(\"https://cdn.plot.ly/plotly-latest.min.js\")","pinned":false,"hidden":false,"mode":"ojs"},{"id":274,"value":"plotIt=async function(trace,layout){\n if(typeof(plotly)==\"undefined\"){\n let plotly= await require(\"https://cdn.plot.ly/plotly-latest.min.js\")\n }\n let div = DOM.element('div')\n trace=trace||false\n return plotly.plot(div,trace,layout)\n}","pinned":true,"hidden":false,"mode":"ojs"},{"id":230,"value":"$31","pinned":true,"hidden":false,"mode":"ojs"},{"id":35,"value":"html`\n

Results

This is a live report so the results sections is assembled is real-time when the reader authenticates in the data platform where the private data is hosted.

Authentication

Since this report has to be authenticated, the first step requires the reader to login the Box account with access to the private reports.

\n`\n","pinned":false,"hidden":false,"mode":"ojs"},{"id":75,"value":"login={\n let div = DOM.element('div')\n let h = ''\n let loggedIn=false\n if(loggedIn){\n h += '

You are logged in as ...

'\n div.innerHTML=h\n }else{\n h += '

You are not logged in, please do so: Login .

'\n div.innerHTML=h\n }\n return div\n}","pinned":false,"hidden":false,"mode":"ojs"},{"id":119,"value":"{\n let msg = html`

Status: login needed

`\n //debugger\n let code = new URLSearchParams(html``.search).get('code')\n if(code){ // we're halfway through the OAUTH dance\n msg = html`

Status: login in progress ... just got the code: ${code}

`\n let xhr = new XMLHttpRequest();\n xhr.open(\"POST\", \"https://api.box.com/oauth2/token\");\n xhr.setRequestHeader(\"cache-control\", \"no-cache\");\n xhr.setRequestHeader(\"credentials\", \"same-origin\");\n //debugger\n xhr.onload=x=>{\n ageDistribution(JSON.parse(x.target.response))\n let li = document.getElementById('loginMsg')\n li.style.color='green'\n li.innerHTML='you are logged in as ...'\n }\n xhr.send(`grant_type=authorization_code&code=${code}&client_id=rq2ab1uuvrzp86oa0yehgjibqf7arxy5&client_secret=NItekKKQEqQBgRsU0qnEBVY3zP0nvieh`);\n } \n return msg\n}","pinned":false,"hidden":false,"mode":"ojs"},{"id":194,"value":"html`

Authenticated Results will show up here after longin...

`\n","pinned":false,"hidden":false,"mode":"ojs"}],"files":"$W32","stdlib":"1","titling":"auto"}