ssm_voting_mps.Rmd

---
title: "Same Sex Marriage Voting Intention of NSW MPs"
author: "Neil Saunders"
date: "compiled `r Sys.time()`"
output:
  html_document:
    keep_md: yes
    number_sections: yes
    toc: yes
    toc_float: yes
  github_document:
    toc: yes
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE, warning = FALSE, message = FALSE)
library(rvest)
library(rgdal)
library(leaflet)
```

# Introduction
This document shows how data can be added to a map using the R [leaflet package](https://rstudio.github.io/leaflet/). In this example we obtain from an online news article data describing the voting intentions of Australian politicians with regard to same sex marriage law and map that onto a map of New South Wales.

# Data
## NSW electorates
The map of NSW electorates is obtained as a shapefile [from the Australian ELectoral Commission](http://www.aec.gov.au/Electorates/gis/gis_datadownload.htm).

```{r nsw-map}
nsw <- readOGR("../../data//nsw-esri-06042016", "NSW_electoral_boundaries_25-02-2016")
```

## Voting intention
Voting intentions of MPs are obtained from a HTML table in [this ABC News article](http://www.abc.net.au/news/2017-11-14/same-sex-marriage-if-the-survey-says-yes-how-will-your-mp-vote/9104112).

```{r mp-votes}
abc <- read_html("http://www.abc.net.au/news/2017-11-14/same-sex-marriage-if-the-survey-says-yes-how-will-your-mp-vote/9104112")

ssm_votes_mp <- abc %>% 
  html_nodes("table") %>% 
  .[[1]] %>% 
  html_table()
```

# Joining data to the map
The process of joining data is as follows:

- match electoral division names in the shapefile to those in the data frame from the news article
- use the matches to add voting intention to the `SpatialPolygonsDataFrame`
- simplify the voting intentions by collapsing the "it's complicated" and "Yes" variations

```{r join-data}
m1 <- match(nsw$Elect_div, ssm_votes_mp$Electorate)

nsw$Vote <- ssm_votes_mp$`How they would vote`[m1]
nsw$Vote <- gsub("complicated .+$", "complicated", nsw$Vote)
nsw$Vote <- gsub("Yes .+$", "Yes", nsw$Vote)
```

# Map generation
The voting data can then be added to the map using `leaflet`. Note that two electorates (New England and Bennelong) are missing as they do not currently have a representative.

Mouse-over each electorate indicates the name of the politician and their detailed voting intention.

```{r map-data}
factpal <- colorFactor("Spectral", nsw$Vote)

leaflet(nsw) %>% 
  addPolygons(fillColor = ~factpal(Vote), 
              label = paste(ssm_votes_mp$MP[m1], ssm_votes_mp$`How they would vote`[m1], sep = " : "), 
              fillOpacity = 1, 
              weight = 2) %>% 
  addLegend(position = "bottomright", 
            pal = factpal, 
            values = ~Vote, 
            opacity = 1)
```