---
title: "Analyse de l'incidence du syndrôme grippal"
author: "Olivier"
output:
  pdf_document:
    toc: true
  html_document:
    toc: true
    theme: journal
documentclass: article
classoption: a4paper
header-includes:
- \usepackage[french]{babel}
- \usepackage[upright]{fourier}
- \hypersetup{colorlinks=true,pagebackref=true}
---

NB: j'ai repris le header Konrad Kinsen

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

## Import des données

```{r}
data_url <- "https://www.sentiweb.fr/datasets/incidence-PAY-3.csv"

seasonal_flu <- read.csv(data_url, 
                         skip = 1,
                         na.strings = "_")

#autre option 
#seasonnal_flu <read.csv2("incidence-PAY-3.csv", skip = 1, na.strings = "_")


class(seasonal_flu$week)
class(seasonal_flu$inc)
```

Chargement d'un paquet qui prend en compte le format de date.

```{r}
library(parsedate)
```

Puis transformation:

```{r}

convert_date <- function(x){
    ws = paste(x)
    iso = paste0(substring(ws, 1 , 4),
           "-W",
           substring(ws, 5 ,6))
    as.character(parsedate::parse_iso_8601(iso))
}

seasonal_flu$date <- as.Date(sapply(seasonal_flu$week, convert_date))

class(seasonal_flu$week)

```

### ordering data 

```{r}
seasonal_flu <- seasonal_flu[order(seasonal_flu$date),]
```


Verification visuelle:

```{r}
head(seasonal_flu)
```

Vérification si toutes les semaines font bien 7j.

```{r}
all(diff(seasonal_flu$date) == 7)
```

Verfication à l'aide d'un graphique:

```{r}
with(seasonal_flu, plot(date, inc, type = "l"))
```

Si on regarde les 200 dernières senaines (effet COVID):

```{r}
with(tail(seasonal_flu, 200), plot(date, inc, type = "l"))
```

## exclusion ce certaine données



```{r}
pic_annule <- function(annee){
    debut = paste0(annee - 1,  "-08-01")
    fin = paste0(annee,  "-08-01")
    semaines = seasonal_flu$date > debut & seasonal_flu$date <= fin
    sum(seasonal_flu$inc[semaines], na.rm = TRUE)
}
```


J'ai gardé les mêmes années que l'instructeur, au début pour avoir une année complête et à la fin pour avoir des résultats similaires. 

```{r}
annees <- 1986:2017
```

```{r}
incidence_annuelle <- data.frame(annee = annees,
                                incidence = sapply(annees, pic_annule))
```

```{r}
head(incidence_annuelle)

plot(incidence_annuelle$incidence, type = "p")
```

```{r}
head(incidence_annuelle[order(-incidence_annuelle$incidence),])
```


```{r}
hist(incidence_annuelle$incidence, breaks = 10)
```