Basic Parliament Plots"
In ggparliament: Parliament Plots

```r

knitr::opts_chunk$set( collapse = TRUE, comment = "#>" )

library(ggparliament) library(dplyr) library(ggplot2) require(tidyr) require(magrittr) source("../R/parliament_data.R") source("../R/geom_parliament_seats.R") source("../R/geom_highlight_government.R") source("../R/helper_funcs.R") source("../R/draw_majoritythreshold.R") source("../R/draw_partylabels.R") source("../R/draw_majoritythreshold.R") source("../R/draw_totalseats.R") source("../R/theme_ggparliament.R") load("../R/sysdata.rda")

## Introduction to `ggparliament`

`ggparliament` is a `ggplot2` extension for creating parliament plots. This `R` package allows users to plot different styles of legislatures. They are:

- semicircle layouts 

- horseshoe parliaments

- opposing bench-style parliaments

- the classroom layout

- a circular parliament.

To plot a parliament in `ggparliament`, we start with a data frame containing the number of seats per political party. This package comes with recent election results from the UK, Australia, the United Sates, Russia, and Germany. `election_data` also provides party colour, abbreviated and full versions of the party name, as well as election returns over several election cycles.

### Semicircle Parliament

To plot a semicircle legislature, run the following commands:

1. Gather some election results.

```r
us_rep <- election_data %>%
    filter(country == "USA" &
               year == 2016 &
               house == "Representatives")
head(us_rep)

Input the data into ggparliament::parliament_data.

This function unravels election results by party to individual seats. Users must define the input data frame, the type of legislature, the number of rows in the legislature, and a numeric variable which expands the number of seats to the total number of members in the legislature.

us_house_semicircle <- parliament_data(election_data = us_rep,
                                       type = "semicircle",
                                       parl_rows = 10,
                                       party_seats = us_rep$seats)
head(us_house_semicircle)

The resulting data frame returns the x and y coordinates of each individual seat for a 10-row semicircle parliament.

Plot the data in ggparliament

The user can take this data frame and plot by the x and y coordinates using geom_parliament_seats().

```r

us <- ggplot(us_house_semicircle, aes(x = x, y = y, colour = party_short)) + geom_parliament_seats() + theme_ggparliament() + labs(colour = NULL, title = "United States Congress") + scale_colour_manual(values = us_house_semicircle$colour, limits = us_house_semicircle$party_short)

Other parliaments are plotted using the same method. The type of parliament should change - i.e, for a horseshoe parliament, specify horseshoe and so on. There are five types - "semicircle", "horseshoe", "opposing_benches", "classroom" and "circle".

### Horsehoe Parliament

```r
australia <- election_data %>%
    filter(country == "Australia" &
               house == "Representatives" &
               year == 2016) 

australia_horseshoe <- parliament_data(election_data = australia,
                                       party_seats = australia$seats,
                                       parl_rows = 4,
                                       type = "horseshoe")

au <- ggplot(australia_horseshoe, aes(x, y, colour = party_short)) +
    geom_parliament_seats() + 
    theme_ggparliament() +
    labs(colour = NULL, 
         title = "Australian Parliament") +
    scale_colour_manual(values = australia$colour, 
                        limits = australia$party_short) + 
    theme(legend.position = 'bottom')

au

Opposing Benches Parliament

The opposing bench style of parliament calls for a variable to split the left and right bench. Using the ggparliament election data, we split on the government column which is a binary variable.

#get the data for the last election
uk_data <- election_data %>%
  filter(country == "UK") %>%
  filter(year == 2017) %>% #parliament_data() can be called in a dplyr chain.
  parliament_data(election_data = ., 
                  party_seats = .$seats,
                  #need to include grouping for opposing benches by definition
                  group = .$government,
                  parl_rows = 12,
                  type = "opposing_benches")

#plot it
uk <- ggplot(uk_data, aes(x, y, colour = party_short)) +
  scale_colour_manual(values = uk_data$colour, limits = uk_data$party_short) +
  geom_parliament_seats() +
  theme_ggparliament()

uk

Classroom parliament

russia_classroom <- election_data %>%
  filter(country == "Russia" &
    house == "Duma" &
    year == 2016) %>% 
  parliament_data(election_data = .,
                  party_seats = .$seats,
                  parl_rows = 11,
                  type = "classroom")

rus <- ggplot(russia_classroom, aes(x, y, colour = party_short)) +
  geom_parliament_seats() +
  theme_ggparliament() +
  labs(
    colour = NULL,
    title = "Russian Duma") +
  scale_colour_manual(
    values = russia_classroom$colour,
    limits = russia_classroom$party_short) +
  theme(legend.position = "bottom")

rus

Circle parliament

There are few national legislative bodies that use a circular parliament. Given that this vignette is primarily to explain how the process works, we are re-using the Russian election returns for this example.

russia_circle <- election_data %>%
  filter(country == "Russia" &
    house == "Duma" &
    year == 2016) %>% 
  parliament_data(election_data = .,
    party_seats = .$seats,
    parl_rows = 11,
    type = "circle")

russia_circle_example <- ggplot(russia_circle, aes(x, y, colour = party_short)) +
  geom_parliament_seats() +
  theme_ggparliament() +
  scale_colour_manual(
    values = russia_circle$colour,
    limits = russia_circle$party_short) +
  labs(colour = NULL) +
  theme(legend.position = "bottom")

russia_circle_example

Though not suggested, data can also be plotted simply from a vector of the parties winning seats as below.

#a made up example of a vector of seats
seats <- c(rep("democrats", 40), rep("republicans", 30), rep("socialists", 20), rep("nationalists", 10))

made_up_layout <- parliament_data(election_data = NULL,
    party_seats = as.numeric(table(seats)),
    parl_rows = 4,
    type = "semicircle")

made_up_layout$party = seats

plot <- ggplot(made_up_layout, aes(x, y, colour = party)) +
    geom_parliament_seats() +
    theme_ggparliament() +
    labs(colour = NULL) +
    theme(legend.position = "bottom")

plot

You can also create a data frame in R and plot your results.

house_of_reps <- data.frame(
  party = c("Vacancies","Republicans", "Democrats", "Vacancies"),
  seats = c(3, 236, 193, 3),
  color = c("gray", "red", "blue", "gray")
) 
house_of_reps <- house_of_reps %>% 
  mutate(party = as.character(party), color = as.character(color)) %>%
  parliament_data(election_data = .,
                            parl_rows = 8,
                            party_seats = .$seats,
                            type = 'semicircle')
ggplot(data = house_of_reps) +
  geom_parliament_seats(aes(x = x,  y = y, color = party)) +
  theme_ggparliament() +
  scale_color_manual(values = house_of_reps$color, 
                     limits = house_of_reps$party)