Skip to content

M
mooc-rr
Commit faa16ada authored by Loick Klpr's avatar Loick Klpr
Browse files
Options

Réalisation de l'exercice évalué par les pairs.

parent 84357140
Show whitespace changes
Inline Side-by-side
Showing with 1118 additions and 36 deletions
.gitignore 0 → 100644
View file @ faa16ada
.Rproj.user
.Rhistory
.RData
.Ruserdata
module3/exo1/analyse-syndrome-grippal.html
View file @ faa16ada
......@@ -471,12 +471,12 @@ summary {
<p>Regardons ce que nous avons obtenu:</p>
<pre class="r"><code>head(data)</code></pre>
<pre><code>## week indicator inc inc_low inc_up inc100 inc100_low inc100_up geo_insee
## 1 202105 3 22491 18436 26546 34 28 40 FR
## 2 202104 3 25804 21491 30117 39 32 46 FR
## 3 202103 3 21810 17894 25726 33 27 39 FR
## 4 202102 3 17320 13906 20734 26 21 31 FR
## 5 202101 3 21799 17778 25820 33 27 39 FR
## 6 202053 3 21220 16498 25942 32 25 39 FR
## 1 202106 3 23816 18867 28765 36 29 43 FR
## 2 202105 3 22491 18436 26546 34 28 40 FR
## 3 202104 3 25804 21491 30117 39 32 46 FR
## 4 202103 3 21810 17894 25726 33 27 39 FR
## 5 202102 3 17320 13906 20734 26 21 31 FR
## 6 202101 3 21799 17778 25820 33 27 39 FR
## geo_name
## 1 France
## 2 France
......@@ -486,26 +486,26 @@ summary {
## 6 France</code></pre>
<pre class="r"><code>tail(data)</code></pre>
<pre><code>## week indicator inc inc_low inc_up inc100 inc100_low inc100_up
## 1888 198449 3 101073 81684 120462 184 149 219
## 1889 198448 3 78620 60634 96606 143 110 176
## 1890 198447 3 72029 54274 89784 131 99 163
## 1891 198446 3 87330 67686 106974 159 123 195
## 1892 198445 3 135223 101414 169032 246 184 308
## 1893 198444 3 68422 20056 116788 125 37 213
## 1889 198449 3 101073 81684 120462 184 149 219
## 1890 198448 3 78620 60634 96606 143 110 176
## 1891 198447 3 72029 54274 89784 131 99 163
## 1892 198446 3 87330 67686 106974 159 123 195
## 1893 198445 3 135223 101414 169032 246 184 308
## 1894 198444 3 68422 20056 116788 125 37 213
## geo_insee geo_name
## 1888 FR France
## 1889 FR France
## 1890 FR France
## 1891 FR France
## 1892 FR France
## 1893 FR France</code></pre>
## 1893 FR France
## 1894 FR France</code></pre>
<p>Y a-t-il des points manquants dans nos données ?</p>
<pre class="r"><code>na_records = apply(data, 1, function (x) any(is.na(x)))
data[na_records,]</code></pre>
<pre><code>## week indicator inc inc_low inc_up inc100 inc100_low inc100_up geo_insee
## 1657 198919 3 0 NA NA 0 NA NA FR
## 1658 198919 3 0 NA NA 0 NA NA FR
## geo_name
## 1657 France</code></pre>
## 1658 France</code></pre>
<p>Les deux colonnes qui nous intéressent sont <code>week</code> et <code>inc</code>. Vérifions leurs classes:</p>
<pre class="r"><code>class(data$week)</code></pre>
<pre><code>## [1] &quot;integer&quot;</code></pre>
......@@ -537,10 +537,10 @@ data[na_records,]</code></pre>
<h3>Inspection</h3>
<p>Regardons enfin à quoi ressemblent nos données !</p>
<pre class="r"><code>plot(data$date, data$inc, type=&quot;l&quot;, xlab=&quot;Date&quot;, ylab=&quot;Incidence hebdomadaire&quot;)</code></pre>
<p><img src="data:image/png;base64,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" width="672" /></p>
<p><img src="data:image/png;base64,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" width="672" /></p>
<p>Un zoom sur les dernières années montre mieux la localisation des pics en hiver. Le creux des incidences se trouve en été.</p>
<pre class="r"><code>with(tail(data, 200), plot(date, inc, type=&quot;l&quot;, xlab=&quot;Date&quot;, ylab=&quot;Incidence hebdomadaire&quot;))</code></pre>
<p><img src="data:image/png;base64,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" width="672" /></p>
<p><img src="data:image/png;base64,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" width="672" /></p>
</div>
</div>
<div id="lincidence-annuelle" class="section level2">
......
module3/exo2/exercice_fr.Rmd
View file @ faa16ada
......@@ -13,7 +13,6 @@ knitr::opts_chunk$set(echo = TRUE)
Data url web et local pour les données de la varicelle
```{r}
data_url = "https://www.sentiweb.fr/datasets/incidence-PAY-7.csv"
data_url_local = "D:/PhD/formation recherche reproductible/mooc-rr/module3/exo2/incidence-PAY-7.csv"
```
......@@ -25,6 +24,150 @@ data_locale = read.csv(data_url_local, skip = 1)
Plot les
```{r}
plot(data_locale$indicator ~ data_locale$week)
plot(data_locale$inc ~ data_locale$week,type = "l",col = "blue",xlab = "week", ylab = "incidence")
```
## Préparation des données
Les données de l'incidence de la varicelle sont disponibles du site Web du [Réseau Sentinelles](http://www.sentiweb.fr/). Nous les récupérons sous forme d'un fichier en format CSV dont chaque ligne correspond à une semaine de la période demandée. Nous téléchargeons toujours le jeu de données complet, qui commence en 1990 et se termine avec une semaine récente. L'URL est:
```{r}
data_url = "https://www.sentiweb.fr/datasets/incidence-PAY-7.csv"
```
Voici l'explication des colonnes donnée sur le [sur le site d'origine](https://ns.sentiweb.fr/incidence/csv-schema-v1.json):
| Nom de colonne | Libellé de colonne |
|----------------+-----------------------------------------------------------------------------------------------------------------------------------|
| `week` | Semaine calendaire (ISO 8601) |
| `indicator` | Code de l'indicateur de surveillance |
| `inc` | Estimation de l'incidence de consultations en nombre de cas |
| `inc_low` | Estimation de la borne inférieure de l'IC95% du nombre de cas de consultation |
| `inc_up` | Estimation de la borne supérieure de l'IC95% du nombre de cas de consultation |
| `inc100` | Estimation du taux d'incidence du nombre de cas de consultation (en cas pour 100,000 habitants) |
| `inc100_low` | Estimation de la borne inférieure de l'IC95% du taux d'incidence du nombre de cas de consultation (en cas pour 100,000 habitants) |
| `inc100_up` | Estimation de la borne supérieure de l'IC95% du taux d'incidence du nombre de cas de consultation (en cas pour 100,000 habitants) |
| `geo_insee` | Code de la zone géographique concernée (Code INSEE) http://www.insee.fr/fr/methodes/nomenclatures/cog/ |
| `geo_name` | Libellé de la zone géographique (ce libellé peut être modifié sans préavis) |
La première ligne du fichier CSV est un commentaire, que nous ignorons en précisant `skip=1`.
### Téléchargement
```{r}
data = read.csv(data_url_local, skip=1)
```
Regardons ce que nous avons obtenu:
```{r}
head(data)
tail(data)
```
Y a-t-il des points manquants dans nos données ?
```{r}
na_records = apply(data, 1, function (x) any(is.na(x)))
data[na_records,]
```
Non pas de données manquantes.
Les deux colonnes qui nous intéressent sont `week` et `inc`. Vérifions leurs classes:
```{r}
class(data$week)
class(data$inc)
```
Ce sont des entiers, tout va bien !
### Conversion des numéros de semaine
La gestion des dates est toujours un sujet délicat. Il y a un grand nombre de conventions différentes qu'il ne faut pas confondre. Notre jeux de données utilise un format que peu de logiciels savent traiter: les semaines en format [ISO-8601](https://en.wikipedia.org/wiki/ISO_8601). En `R`, il est géré par la bibliothèque [parsedate](https://cran.r-project.org/package=parsedate):
```{r}
library(parsedate)
```
Pour faciliter le traitement suivant, nous remplaçons ces semaines par les dates qui correspondent aux lundis. Voici une petite fonction qui fait la conversion pour une seule valeur:
```{r}
convert_week = function(w) {
ws = paste(w)
iso = paste0(substring(ws, 1, 4), "-W", substring(ws, 5, 6))
as.character(parse_iso_8601(iso))
}
```
Nous appliquons cette fonction à tous les points, créant une nouvelle colonne `date` dans notre jeu de données:
```{r}
data$date = as.Date(convert_week(data$week))
```
Vérifions qu'elle est de classe `Date`:
```{r}
class(data$date)
```
Les points sont dans l'ordre chronologique inverse, il est donc utile de les trier:
```{r}
data = data[order(data$date),]
```
C'est l'occasion pour faire une vérification: nos dates doivent être séparées d'exactement sept jours:
```{r}
all(diff(data$date) == 7)
```
### Inspection
Regardons enfin à quoi ressemblent nos données !
```{r}
plot(data$date, data$inc, type="l", xlab="Date", ylab="Incidence hebdomadaire")
```
Un zoom sur les dernières années montre mieux la localisation des pics en hiver. Le creux des incidences se trouve en été.
```{r}
with(tail(data, 200), plot(date, inc, type="l", xlab="Date", ylab="Incidence hebdomadaire"))
```
## L'incidence annuelle
### Calcul
Étant donné que le pic de l'épidémie se situe en hiver, à cheval entre deux années civiles, nous définissons la période de référence entre deux minima de l'incidence, du 1er septembre de l'année $N$ au 1er septembre de l'année $N+1$. Nous mettons l'année $N+1$ comme étiquette sur cette année décalée, car le pic de l'épidémie est toujours au début de l'année $N+1$. Comme l'incidence de varicelle est très faible en été, cette modification ne risque pas de fausser nos conclusions.
L'argument `na.rm=True` dans la sommation précise qu'il faut supprimer les points manquants.
```{r}
pic_annuel = function(annee) {
debut = paste0(annee-1,"-09-01")
fin = paste0(annee,"-09-01")
semaines = data$date > debut & data$date <= fin
sum(data$inc[semaines], na.rm=TRUE)
}
```
Nous devons aussi faire attention aux premières et dernières années de notre jeux de données. Les données commencent en décembre 1990, ce qui ne permet pas de quantifier complètement le pic attribué à 1991 Nous l'enlevons donc de notre analyse. Par contre, pour une exécution en octobre 2020, les données se terminent après le 1er août 2020, ce qui nous permet d'inclure cette année.
```{r}
annees = 1992:2019
```
Nous créons un nouveau jeu de données pour l'incidence annuelle, en applicant la fonction `pic_annuel` à chaque année:
```{r}
inc_annuelle = data.frame(annee = annees,
incidence = sapply(annees, pic_annuel))
head(inc_annuelle)
```
### Inspection
Voici les incidences annuelles en graphique:
```{r}
plot(inc_annuelle, type="p", xlab="Année", ylab="Incidence annuelle")
```
### Identification des épidémies les plus fortes
Une liste triée par ordre décroissant d'incidence annuelle permet de plus facilement repérer les valeurs les plus élevées:
```{r}
head(inc_annuelle[order(+inc_annuelle$incidence),])
```
Enfin, un histogramme montre bien que les épidémies fortes, qui touchent environ 10% de la population française, sont assez rares: il y en eu trois au cours des 35 dernières années.
```{r}
hist(inc_annuelle$incidence, breaks=10, xlab="Incidence annuelle", ylab="Nb d'observations", main="")
```
module3/exo2/exercice_fr.html
View file @ faa16ada
This source diff could not be displayed because it is too large. You can view the blob instead.
module3/exo3/exercice_fr.Rmd
View file @ faa16ada
---
title: "Votre titre"
author: "Votre nom"
date: "La date du jour"
title: "TP évalué par les pairs"
author: "Loïck Kléparski"
date: "18/02/2021"
output: html_document
---
......@@ -10,24 +10,187 @@ output: html_document
knitr::opts_chunk$set(echo = TRUE)
```
## Quelques explications
## Etude de la concentration de CO2 dans l'atmosphère depuis 1958
Ceci est un document R markdown que vous pouvez aisément exporter au format HTML, PDF, et MS Word. Pour plus de détails sur R Markdown consultez <http://rmarkdown.rstudio.com>.
Fichier télécharger le 18/02/2021
Lorsque vous cliquerez sur le bouton **Knit** ce document sera compilé afin de ré-exécuter le code R et d'inclure les résultats dans un document final. Comme nous vous l'avons montré dans la vidéo, on inclue du code R de la façon suivante:
Atmospheric CO2 concentrations (ppm) derived from in situ air measurements at Mauna Loa, Observatory, Hawaii: Latitude 19.5°N Longitude 155.6°W Elevation 3397m.
```{r cars}
summary(cars)
Source: R. F. Keeling, S. J. Walker, S. C. Piper and A. F. Bollenbacher. Scripps CO2 Program ( http://scrippsco2.ucsd.edu ). Scripps Institution of Oceanography (SIO). University of California. University of California. La Jolla, California USA 92093-0244.
Données hebdomadaires.
Les analyses effectuées pour décomposer la série temporelle ont été réalisées à partir du livre *Numerical Ecology* écrit par Pierre et Louis Lengendre et édité par Elsevier en 1998.
`url_data_co2_web` = lien pour télécharger les données.
`url_data_co2_local` = lien des données en local.
```{r}
url_data_co2_web = "https://scrippsco2.ucsd.edu/assets/data/atmospheric/stations/in_situ_co2/weekly/weekly_in_situ_co2_mlo.csv"
url_data_co2_locale = "D:/PhD/formation recherche reproductible/mooc-rr/module3/exo3/weekly_in_situ_co2_mlo.csv"
```
### 1. Réalisez un graphique qui vous montrera une oscillation périodique superposée à une évolution systématique plus lente.
Après un examen visiuel du fichier csv, l'on constate que les données commencent à la ligne 45. L'on peut donc "skipper" les 44 premières lignes. La première colonne correspond aux dates et la seconde aux concentration de CO2 en ppm.
```{r}
data_co2 = read.csv(url_data_co2_locale, skip = 44, col.names = c("dates","CO2"), header = FALSE)
```
On vérifie le type de données dans chaque colonne.
```{r}
class(data_co2$dates)
class(data_co2$CO2)
```
Il faut convertir le vecteur *dates* au format date avec la fonction `as.Date`. Les dates renseignées étant déjà au bon format, cela ne devrait pas poser de problème. Les résultats sont placés dans un vecteur *dates2*:
```{r}
data_co2$dates2 = as.Date(data_co2$dates, format = "%Y-%m-%d")
```
On vérifie que cela a fonctionné:
```{r}
class(data_co2$dates2)
```
Y a-t-il des points manquants dans nos données ?
```{r}
na_records = apply(data_co2, 1, function (x) any(is.na(x)))
data_co2[na_records,]
```
On peut à présent représenter graphiquement la relation entre la concentration en CO2 et le temps:
```{r}
plot(data_co2$CO2 ~ data_co2$dates2, type = "l", col = "blue", xlab = "Time", ylab = "CO2 in ppm")
```
On zoom sur la fin des données pour mieux voir la périodicité.
```{r}
plot(data_co2$CO2[c(3000:3208)] ~ data_co2$dates2[c(3000:3208)], type = "l", col = "blue", xlab = "Time", ylab = "CO2 in ppm")
```
On peut donc constater la présence d'une périodicité, qui semble à première vue être annuelle, et d'une tendance générale à l'augmentation.
### 2. Séparez ces deux phénomènes. Caractérisez l'oscillation périodique. Proposez un modèle simple de la contribution lente, estimez ses paramètres et tentez une extrapolation jusqu'à 2025 (dans le but de pouvoir valider le modèle par des observations futures).
Pour décomposer cette série, deux modèles de décomposition s'offrent à nous: un modèle additif ou multiplicatif. La périodicité étant annuelle, nous allons estimer les moyennes et les écarts types annuels. Si aucune relation n'apparait entre ces deux paramètres, on appliquera un modèle additif pour décomposer la série, sinon on appliquera un modèle multiplicatif.
On charge le package `lubridate` qui va nous permettre d'utiliser la fonction `year` que nous allons utiliser pour isoler les concentrations en CO2 qui appartiennent toutes à la même année.
```{r include=FALSE}
library(lubridate)
```
Et on peut aussi aisément inclure des figures. Par exemple:
On créé deux vecteurs vides, `moyenne_ann` et `et_ann` pour stocker la concentration en CO2 moyenne de chaque année et l'écart type associé
```{r}
moyenne_ann = rep(NA, length(min(year(data_co2$dates2)):max(year(data_co2$dates2))))
et_ann = rep(NA, length(min(year(data_co2$dates2)):max(year(data_co2$dates2))))
```
```{r pressure, echo=FALSE}
plot(pressure)
On effectue ensuite une boucle de 1958 (année minimale (`min(year(data_co2$dates2))`)) à 2021 (année maximale (`max(year(data_co2$dates2))`)). On définit un conteur `a` égal à 1 dont la valeur changera à chaque itération pour nous permettre de stocker les moyennes et les écarts types à la bonne position dans nos vecteurs vides `moyenne_ann` et `et_ann`. On définit également un vecteur *logique* `test` qui va nous permettre d'isoler dans la matrice `data_co2$CO2` les concentration en CO2 correspondant à l'année de l'itérations.
```{r}
a=1
for (i in min(year(data_co2$dates2)):max(year(data_co2$dates2))){
test = (year(data_co2$dates2) == i) == TRUE
moyenne_ann[a] = mean(data_co2$CO2[test == TRUE])
et_ann[a] = sd(data_co2$CO2[test == TRUE])
a=a+1
}
```
Vous remarquerez le paramètre `echo = FALSE` qui indique que le code ne doit pas apparaître dans la version finale du document. Nous vous recommandons dans le cadre de ce MOOC de ne pas utiliser ce paramètre car l'objectif est que vos analyses de données soient parfaitement transparentes pour être reproductibles.
On peut à présent plotter les résultats:
```{r}
plot(moyenne_ann ~ et_ann, xlab = "annual mean", ylab = "standard deviation")
```
Comme les résultats ne sont pas stockés dans les fichiers Rmd, pour faciliter la relecture de vos analyses par d'autres personnes, vous aurez donc intérêt à générer un HTML ou un PDF et à le commiter.
Et calculer le coefficient de corrélation de Pearson `r` entre les deux paramètres avec la fonction `cor`.
```{r}
r = cor(moyenne_ann, et_ann, method = "pearson")
r
```
Il n'y a pas de relation entre les deux variables, on peut donc utiliser un modèle additif pour décomposer la série.
Pour extraire la tendance des données, on applique une moyenne mobile simple d'ordre n = 52 (nombre de semaines dans une année) avec la fonction `movavg` du package `pracma`
```{r}
library(pracma)
trend = movavg(data_co2$CO2, 52, type = "s")
```
On plot le résultat
```{r}
plot(trend ~ data_co2$dates2, type = "l", col = "red", xlab = "", ylab = "", axes = FALSE)
par(new = T)
plot(data_co2$CO2 ~ data_co2$dates2, type = "l", col = "blue", xlab = "Time", ylab = "CO2 in ppm")
legend(1, 400, legend = c("tendance", "raw data"), col = c("red", "blue"), lty = 1, cex = 1)
```
On extrait ensuite la tendance des données avec la formule suivante $Y_{i Detrend} = Y_{i} - \hat{Y_{i}}$, avec $Y_{i}$ les concentrations en CO2 et $\hat{Y_{i}}$ la tendance estimé avec la moyenne mobile.
```{r}
data_co2$CO2dt = data_co2$CO2 - trend
```
On plot le résultat pour voir si cela a marché.
```{r}
plot(data_co2$CO2dt ~ data_co2$dates2, type = "l", col = "blue", xlab = "Time", ylab = "Detrend CO2 in ppm")
```
Pour caractériser l'oscillation périodique, nous allons effectué une mesure d'autocorrelation avec la fonction `acf` sur les données détentancées:
```{r}
auto_cor = acf(data_co2$CO2dt, type = "correlation", lag.max = 100)
```
On observe un décalage d'environs 52-53 pas de temps, ce qui correspond à la durée d'une année en semaine. Nous avons donc une périodicité annuelle.
On va ensuite ajuster un polynome de degré 3 avec la fonction `polyfit` du package `pracma`.
```{r}
model = polyfit(seq(1,3208), data_co2$CO2, 3)
```
Voici les paramètres du modèle:
```{r}
model
```
Et sa représentation graphique.
```{r}
output = polyval(model, seq(1,3208))
plot(output ~ data_co2$dates2, type = "l", col = "red", axes = FALSE, xlab = "", ylab = "", lwd = 2)
par(new = T)
plot(data_co2$CO2 ~ data_co2$dates2, type = "l", col = "blue", xlab = "Time", ylab = "CO2 in ppm")
legend(1, 400, legend = c("model", "raw data"), col = c("red", "blue"), lty = 1, cex = 1)
```
Malheureusement, nous ne sommes pas parvenu à ajuster le polynome en considérant les dates en tant que tel.
La dernière valeur de CO2 enregistrée correspond à la date:
```{r}
last_date = data_co2$dates2[length(data_co2$dates2)]
```
Pour extrapoler jusqu'en 2025, il nous faut estimer le nombre de semaine entre le 30 janvier 2021 et le 30 janvier 2025.
```{r}
date1 = as.Date("2021-01-30")
date2 = as.Date("2025-01-30")
difftime(date2, date1, units="weeks")
```
En arrondissant à 209, on peut relancer notre code précédent pour extrapoler jusqu'en 2025:
```{r}
output2 = polyval(model, seq(1,3208+209))
```
Avec le package `lubridate` et la fonction `weeks` l'on va pouvoir compléter notre vecteur `dates2` avec les dates manquantes entre 30-01-2021 et le 30-01-2025.
```{r}
library(lubridate)
date_new = c(data_co2$dates2,last_date+weeks(1:209))
last_date+weeks(1:209)
```
Et représenter graphiquement le résultat.
```{r}
plot(output2 ~ date_new, type = "l", col = "red", xlab = "", ylab = "", lwd = 2)
par(new = T)
plot(c(data_co2$CO2, rep(NA, length(last_date+weeks(1:209)))) ~ date_new, type = "l", col = "blue", axes = FALSE, xlab = "Time", ylab = "CO2 in ppm")
legend(1, 400, legend = c("model", "raw data"), col = c("red", "blue"), lty = 1, cex = 1)
```
Maintenant, à vous de jouer! Vous pouvez effacer toutes ces informations et les remplacer par votre document computationnel.
Notre modèle sous estime beaucoup l'accroissement de la concentration en CO2.
module3/exo3/exercice_fr.html 0 → 100644
View file @ faa16ada
This source diff could not be displayed because it is too large. You can view the blob instead.
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment