##########################################################################################
###Data preparation script for the extraction of government positions from the Manifesto data


### 0. Libraires
require(foreign)

### 1. The original datasets
##  1.1 Get tue Manifesto Project data from https://manifesto-project.wzb.eu/ 
##  and load it. The version of the file used in the script is 2009/10/11
manifestos <- read.csv("./original data/MPDataset_full.csv", as.is=TRUE)

##  1.2 Get the government and party composition files from ParlGov https://manifesto-project.wzb.eu/
##  and load then. The versions of the files used in the script are 26 July 2011
parties <- read.csv("./original data/view_party.csv", as.is=TRUE)
cabinets.temp <- read.csv("./original data/view_cabinet.csv", as.is=TRUE)

### 2. Data manipulation
##  2.1 Merge the manifesto and party files
manipart<-merge(parties, manifestos, by.x="cmp", by.y="party")

## 2.2 Run the correction dates
## Some of the election dates in the Manifesto and ParGov files differ. The source files alignes the dates.
## It is possible that in future versions of the files, the corrections would not be needed.
source('corrections dates cabinets.r')

## 2.3 Create id variables based on parties and elections
cabinets.temp$new_id<-paste(cabinets.temp$party_id, cabinets.temp$election_date)
manipart$edate2<-as.Date(manipart$edate, "%m/%d/%Y")
manipart$new_id<-paste(manipart$party_id, manipart$edate2)

## 2.4 Merge the cabinet, party and manifesto data
cabinets<-merge(cabinets.temp, manipart, by.x="new_id", by.y="new_id")

## 2.5 Create a new variable for ease
cabinets$seats_share <- (cabinets$seats / cabinets$election_seats_total) * 100

## 2.5 Save the table
write.table (cabinets, 'cabinets.txt')

## You can load the data by the following (might need to specify the location of the file)
cabinets<-read.table ('cabinets.txt', as.is=TRUE)