## ----setup, echo = TRUE, include=FALSE-----------------------------------
knitr::opts_chunk$set(echo = FALSE)


## ----cars, echo = TRUE, eval = FALSE-------------------------------------
## install.packages("tidyverse")
## install.packages("dplyr")
## 


## ---- eval = FALSE, echo = TRUE------------------------------------------
## names(data) <- c("new_name", "another_new_name")
## colnames(data)[colnames(data)=="old_name"] <- "new_name"
## 


## ---- echo = TRUE, eval = FALSE------------------------------------------
## rename(data, new_name = old_name)
## select(data, variable = starts_with('S')) #rename
## #columns in a group


## ---- echo = TRUE, eval = FALSE------------------------------------------
## data$Variable[data$Variable== 1] <- "Value"
## data$Variable[data$Variable > 1] <- "Value2"
## 
## recode(Variable, `1` = "Value", .default = "Value2") #dplyr


## replace Variable = "Value" if Variable == 1

## replace Variable = "Value2" if Variable > 1

## 

## ---- echo = TRUE, eval = FALSE------------------------------------------
## data$Variable[data$Variable < 3 &
##             data$Variable2 == "True"] <- NA
## 


## ---- echo = TRUE, eval = FALSE------------------------------------------
## # create new dataset without missing data
## newdata <- na.omit(data)
## 
## # list rows of data that have missing values
## mydata[!complete.cases(mydata),]


## ---- echo = TRUE, eval = FALSE------------------------------------------
## data$Var_log <- log(data$Variable)


## gen Var_log = log(Variable)


## ---- echo = TRUE, eval = FALSE------------------------------------------
## newdata <- mydata[!mpg]


## drop if mpg > 21

## drop mpg


## ---- echo = FALSE, message=FALSE, eval = TRUE---------------------------
rm(list=ls())
library(dplyr)
library(tidyverse)
setwd("~/Dropbox/Workshops/Teaching R/")


## ---- echo = TRUE, message=FALSE, eval = TRUE----------------------------
library(haven)
cses_work <- read_dta("cses_work.dta")


## ---- echo = TRUE, message=FALSE, eval = TRUE----------------------------
ncol(cses_work)


## ---- echo = TRUE, message=FALSE, eval = TRUE----------------------------
cses_org <- cses_work %>%
  dplyr::select(Country = IMD1006_NAM,
                Country_year = IMD1004, 
                Year  = IMD1008_YEAR,
                Age = IMD2001_1, #in years
                Female = IMD2002, #1 is male, 2 is female 
                Education = IMD2003, 
                Income = IMD2006,
                Urban = IMD2007,
                Turnout = IMD3001, 
                Ideology = IMD3006, 
                Identifier =  IMD3005_1,
  )%>%
  drop_na(Turnout, Ideology, Age)



## ---- echo = TRUE, message=FALSE, eval = FALSE---------------------------
## head(cses_org)


## ---- echo = FALSE, message=FALSE, eval = TRUE---------------------------
library(kableExtra)
kable(head(cses_org[,1:6]))%>%
kable_styling(position = "center")


## ---- echo = TRUE, message=FALSE, eval = FALSE---------------------------
## tail(cses_org)

## ---- echo = FALSE, message=FALSE, eval = TRUE---------------------------
library(kableExtra)
kable(tail(cses_org[,1:6]))%>%
kable_styling(position = "center")


## ---- echo = TRUE, message=FALSE, eval = FALSE---------------------------
## str(cses_org)


## ---- echo = TRUE, message=FALSE, eval = FALSE---------------------------
## summary(cses_org)


## ---- eval = FALSE, echo= TRUE-------------------------------------------
## 
## cses_new <- cses_org[which(cses_org$Female==1
## & cses_org$Age > 25), -which(names(cses_org) %in%
##                                c("Ideology", "Identifier"))]


## ---- eval = FALSE, echo= TRUE-------------------------------------------
## cses_new <- subset(cses_org, Female==1 & Age > 25,
##                    select=Country:Turnout))


## ---- echo = TRUE, message=FALSE, eval = TRUE----------------------------
cses_org <- cses_org %>% 
  filter(Female <= 2)%>%
  mutate(Female = (Female - 1))%>%
  filter(Year > 2005)%>%
  filter(!grepl('South', Country)) 


## ---- echo = TRUE, message=FALSE, eval = TRUE----------------------------
unique(cses_work$IMD2002)

unique(cses_org$Female)


## ---- echo = FALSE, message=FALSE, eval = TRUE---------------------------
kable(head(cses_org[,1:6]))%>%
kable_styling(position = "center")


## ---- echo = FALSE, message=FALSE, eval = TRUE---------------------------
kable(tail(cses_org[,1:5]))%>%
kable_styling(position = "center")


## ----echo = TRUE, eval = TRUE--------------------------------------------
vdem_work <- read_dta("vdem_work.dta")



## ---- echo = TRUE, eval = TRUE-------------------------------------------
cses_merge <- merge(cses_org, vdem_work, 
                    by = c("Country", "Year"), 
                    type = "left", match = "all") 



## ---- echo = FALSE, eval = TRUE------------------------------------------

kable(head(cses_merge[,c(1:2,4:5,12,15)]))%>%
kable_styling(position = "center")






## ---- echo = TRUE--------------------------------------------------------

unique(cses_merge$Turnout)


## ---- echo = TRUE, eval = TRUE-------------------------------------------

cses_merge$Turnout[cses_merge$Turnout > 1 ] <- NA




## ---- echo = TRUE, eval = TRUE-------------------------------------------

cses_summary <- cses_merge %>%
  group_by(Country)%>%
  summarize_at(vars(Turnout.mean = "Turnout",
                    Female.mean = "Female"),
               mean, na.rm = TRUE)





## ---- echo = FALSE, eval = TRUE------------------------------------------

kable(cses_summary[c(1:6, 42:45),])%>%
kable_styling(position = "center")




## ---- echo = FALSE, eval = TRUE, message=FALSE---------------------------
library(reshape)


## ---- echo = TRUE, eval = TRUE, message=FALSE, warning=FALSE-------------
melt.data <- melt(cses_merge, id=c("Country","Year"))


## ---- echo = FALSE, eval = TRUE------------------------------------------
kable(melt.data[c(1, 5000, 15000),])%>%
kable_styling(position = "center")



## ---- echo = TRUE, eval = FALSE------------------------------------------
## # cast(data, formula, function)
## Country.means <- cast(melt.data, Country~variable, mean)
## Year.means <- cast(melt.data, Year~variable, mean)
## 


## ---- eval = FALSE, echo = FALSE-----------------------------------------
## 
## library(knitr)
## 
## purl("~/Dropbox/Workshops/Teaching R/Data Managment R.rmd", output = "~/Dropbox/Workshops/Teaching R/Data_Managment_R_files/data-manage_Jan_20_2020.R")
##