data analyst • mediapart
import { sondages, regions, desc } from '@taniki/regionales2021-donnees-nsppolls'
import { resultats, desc as descBrice } from '@taniki/regionales2021-resultats-premiertour'
regions
regionSondages = sondages.filter(i => i.region == regionSelect)
premierTour = tourSondages(regionSondages, 'Premier tour')
Inputs.table(secondTour)
secondTour = tourSondages(regionSondages, 'Deuxième tour')
tourSondages = (sondages, tour) => {
return sondages
.map(s => [s, s.tours.find(t => t.tour == tour)])
.map(([s, x]) => x?.hypotheses.map(h => ({sondage: s.id, date: s.fin_enquete, n: h.sous_echantillon, ...h})))
.filter(r => r != undefined)
.flat()
}
t1TetesListe = tetesListeRegion('Premier tour')
tetesListeRegion = (tour) => {
let data = [ ...new Set(
tourSondages(regionSondages, tour)
.map(x => x.tetes_liste)
.flat()
.map(t => t.tete_liste)
.filter(t => t != null)
)]
.map(t => [t, candidates.find(c => t.toLowerCase() == `${c['Prénom candidat']} ${c['Nom candidat']}`.toLowerCase() && c['Tête de Liste'] == 'Oui')])
.filter(t => t[1] != undefined )
.map(t => ({ 'name': t[0], info: t[1] }))
return data
}
t2TetesListe = tetesListeRegion('Deuxième tour')
// intentions = (sondages) =>
t1Intentions = premierTour
.map(t => t.tetes_liste.map(tl => ({sondage: t.sondage, date: t.date, n: t.sous_echantillon, hypothese: t.hypothese, ...tl})))
.flat()
.filter(t => t1TetesListe.map(t => t.name).includes(t.tete_liste))
.sort((a,b) => b.date.localeCompare(a.date))
t1Moyennes = [ ...d3.group(t1Intentions, i => i['tete_liste'])]
.map(t => {
let n = d3.sum(t[1], i => i.n)
let p = d3.sum(t[1], i => i.intentions * i.n) / n
let Z = jStat.normal.inv((1+0.95)/2, 0, 1)
let e = Z * Math.sqrt(((p/100) * (1 - (p/100))) / n)
return {
name: t[0],
mean: p,
n,
inf: d3.sum(t[1], i => i.erreur_inf * i.n) / d3.sum(t[1], i => i.n),
sup: d3.sum(t[1], i => i.erreur_sup * i.n) / d3.sum(t[1], i => i.n),
inf2: p - e*100,
sup2: p + e*100,
}
})
Inputs.table(t2Intentions)
t2Intentions = secondTour
.map(t => t.tetes_liste.map(tl => ({sondage: t.sondage, date: t.date, n: t.sous_echantillon, hypothese: t.hypothese, ...tl})))
.flat()
.filter(t => t2TetesListe.map(t => t.name).includes(t.tete_liste))
.sort((a,b) => b.date.localeCompare(a.date))
Inputs.table(t2Moyennes)
t2Moyennes = [ ...d3.group(t2Intentions, i => i['tete_liste'])]
.map(t => {
let n = d3.sum(t[1], i => i.n)
let p = d3.sum(t[1], i => i.intentions * i.n) / n
let Z = jStat.normal.inv((1+0.95)/2, 0, 1)
let e = Z * Math.sqrt(((p/100) * (1 - (p/100))) / n)
return {
name: t[0],
mean: p,
n,
inf: d3.sum(t[1], i => i.erreur_inf * i.n) / d3.sum(t[1], i => i.n),
sup: d3.sum(t[1], i => i.erreur_sup * i.n) / d3.sum(t[1], i => i.n),
inf2: p - e*100,
sup2: p + e*100,
}
})
import { raw as candidates, desc as descMinInt } from '@taniki/regionales2021-donnees-minint'
nuance = (c, name = 'tete_liste') => {
let found = candidates.find(x => `${x['Prénom candidat']} ${x['Nom candidat']}`.toLowerCase() == c[name].toLowerCase())
if (found) return found['Nuance candidat']
else return '???'
}
import {candidateColors, desc as descColors } from '@taniki/regionales2021-colors'
jStat = require('jstat')
One platform to build and deploy the best data apps
Experiment and prototype by building visualizations in live JavaScript notebooks. Collaborate with your team and decide which concepts to build out.
Use Observable Framework to build data apps locally. Use data loaders to build in any language or library, including Python, SQL, and R.
Seamlessly deploy to Observable. Test before you ship, use automatic deploy-on-commit, and ensure your projects are always up-to-date.