README.Rmd

---
output: github_document
---

<!-- README.md is generated from README.Rmd. Please edit that file -->

```{r, include = FALSE}
knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>",
  fig.path = "man/figures/README-",
  out.width = "100%"
)

library(csatimer)
library(ggplot2)
library(tidyverse)


```

# csatimer

`csatimer` provides data on how much airtime French politician receive in the media. 

The datasets contained in the package are also [available](https://github.com/benjaminguinaudeau/csatimer/tree/master/inst/csv) in the csv-format.

## Installation

``` r
# install.packages("devtools")
devtools::install_github("benjaminguinaudeau/csatimer")
```

## Data

*Last Update: February, 2nd 2022* 

Two types of datasets are provided: 

* speaking time of politicians during routine period (no electoral campaigns)
* speaking time of candidates during political campaigns

### Routine period (no electoral campaigns)

Every month, the [CSA](https://www.csa.fr/csapluralisme/tableau) collects how long each political personality speaks on each of the major radio and TV stations. 

This dataset can be accessed using `read_csa_routine` and is continuous between September 2017 and January 2022. 

```{r}
read_csa_routine() %>%
  glimpse
```


It contains the following columns: 

* month <date> `"2017-01-01"`
* station <chr> `"TF1", "France 2", "France Inter", ...`
* prog_type <chr> `"JT", "MAG", "PROG"`
* name <chr> `"Emmanuel Macron", "Marine Le Pen", "François Hollande", ...`
* label <chr> `"Président", "Parti Socialiste", "Ministre", "Les Républicains", ...`
* party <chr> `"Exécutif" , "Ps", "LR", ...`
* time <dbl> 

Speaking times are measured in minutes. For instance, in September 2017, Emanuel Macron spoke 8.4 minutes on TF1.

`prog_type` refers to different types of TV/Radio shows: 

* `JT`: newscast
* `MAG`: political shows
* `PROG`: non-political shows

### Electoral campaigns

During electoral campaigns - presidential, legislative and local - , the CSA publishes airtime of campaigning candidates [data](https://www.csa.fr/Proteger/Garantie-des-droits-et-libertes/Proteger-le-pluralisme-politique/Pendant-une-election) *every two weeks*. 

For each candidate, three different measures are computed: 

* `time_candidat`: speaking time of the candidate themselves
* `time_support`: speaking time of personalities supporting the candidate
* `time_mention`: speaking time of journalists mentioning the candidate

This dataset can be accessed using `read_csa_election("pres_2022")`. 

Following elections are available: 

* Presidential elections 2022: `read_csa_election("pres_2022")`
* Presidential elections 2017: `read_csa_election("pres_2017")`

```{r}
read_csa_election("pres_2022")  %>%
  dplyr::glimpse()
```

### Updating data

Data will be updated as the CSA publishes new routine and campaign data. 

To update the data local version, use `update_csatimer()`

```{r}
# Updating data
update_csatimer()
```

## Example {.tabset}

Here are some graphic examples using `csatimer::read_csa_routine()`

### Example 1

```{r}
read_csa_routine() %>%
  group_by(month) %>%
  summarise(n_hours = sum(time, na.rm = T)/60) %>%
  ggplot(aes(x = month, n_hours)) +
  geom_line() +
  theme_minimal() +
  labs(x = "", y = "Monthly number of hours of political attention", title = "How much do politicians speak in the media?")
```

### Example 2


```{r}
read_csa_routine() %>%
  add_station_dummies %>%
  mutate(private = ifelse(is_private, "Private", "Public")) %>%
  group_by(private, station, month) %>%
  summarise(n_hours = sum(time, na.rm = T)/60) %>%
  ungroup %>%
  mutate(station = fct_reorder(station, n_hours)) %>%
  ggplot(aes(x = station, n_hours, fill = private)) +
  geom_boxplot(outlier.alpha = 0) +
  coord_flip() +
  labs(x = "Media outlet", y = "\nMonthly number of hours of political attention", fill = "") +
  theme_minimal() 
```

### Example 3


```{r}
read_csa_routine() %>% 
  group_by(party, name) %>%
  summarise(n_hours = sum(time, na.rm = T)/60) %>%
  ungroup %>%
  slice_max(n_hours, n = 20) %>%
  mutate(name = fct_reorder(name, n_hours)) %>%
  ggplot(aes(x = name, y = n_hours, fill = party)) +
  geom_col() +
  coord_flip() +
  theme_minimal() +
  labs(x = "Politicians", y = "Total Number of hours")
```

### Example 4


```{r, fig.height = 5, fig.width = 8}
read_csa_routine() %>% 
  filter(month > lubridate::dmy("07-01-2021")) %>%
  filter(station %in% c("TF1", "France 2", "France 3", "C8")) %>%
  group_by(station, party, name) %>%
  summarise(n_hours = sum(time, na.rm = T)/60) %>%
  ungroup %>%
  group_by(station) %>%
  slice_max(n_hours, n = 20) %>%
  ungroup %>%
  mutate(name = fct_reorder(name, n_hours)) %>%
  ggplot(aes(x = name, y = n_hours, fill = party)) +
  geom_col() +
  coord_flip() +
  theme_minimal() +
  facet_wrap(~station, scales = "free") +
  labs(x = "Politicians", y = "Total Number of hours") +
  ggtitle("Which politicians obtained the highest attention on four channel since July 2022?")
```