Capital in the 21st Century: Chapter 3
========================================================

### Data provenance

The data were downloaded as Excel files from: http://piketty.pse.ens.fr/en/capital21c2. 

### Loading relevant libraries and data

### This document depends on the [xlsx](http://cran.r-project.org/web/packages/xlsx/index.html), [reshape2](http://cran.r-project.org/web/packages/reshape2/index.html), [scales](http://cran.r-project.org/web/packages/scales/index.html), and [ggplot2](http://cran.r-project.org/web/packages/ggplot2/index.html) packages.

setwd("/Users/pacha/Dropbox/R/capitalIn21stCenturyinR/Chapter3")

library(XLConnect)
library(WriteXLS)

excel.file <- file.path("../_data/Chapter3TablesFigures.xlsx")

```{r loadCh0,message=FALSE}
library(ggplot2)
library(xlsx)
library(reshape2)
### scales for percent of the axes
library(scales)
```

## Figures 3.1 and 3.2

These figures are based on Table TS3.1 and TS3.2. Here we read
data from the excel file and name the data series.

```{r tabTS3.1, message=FALSE, cache=TRUE}
## Table TS3.1
tab = read.xlsx("../_data/Chapter3TablesFigures.xlsx",sheetName="TS3.1",rowIndex=9:19,colIndex=1:10,header=FALSE)
names(tab) =  c("year","nat","nat_land", "nat_house", "nat_dom", "nat_foreign", "public", "public_ass", "public_debt", "private")
```

```{r tabTS3.2,  message=FALSE, cache=TRUE}
## Table TS3.1
tab32 = read.xlsx("../_data/Chapter3TablesFigures.xlsx",sheetName="TS3.2",rowIndex=9:19,colIndex=1:10,header=FALSE)
names(tab32) =  c("year","nat","nat_land", "nat_house", "nat_dom", "nat_foreign", "public", "public_ass", "public_debt", "private")
```

The caption of the table TS3.1 lists the sources for this table
as 

> Source: Piketty-Zucman 2013, UK.xls, links frozen on 01-24-2013.   				

and 
> Source: Piketty-Zucman 2013, France.xls, links frozen on 01-24-2013

for table TS3.2
Let's make a function that will plot our data:
```{r func3_1, fig.width=8, fig.height=4}
run31_32 = function(tab, num, country){
  fdat <- tab[,c("year","nat_land", "nat_house", "nat_dom", "nat_foreign")]
  names(fdat) <- c("year","Agricultural land","Housing", "Other Domestic Capital", "Net foreign capital")
  
  # Make the data long for stacking
  fdat <- melt(fdat, id.var="year")
  names(fdat) <- c("year","capital_type","percent")
  
  qplot(year, percent, data=fdat, geom="ribbon", position = "stack", 
        fill = capital_type, xlab= "Year", ylab= "Value of National Capital (% National Income)") + scale_y_continuous(labels = percent, limits = c(0, 8)) + 
    ggtitle(paste0("Figure ", num, ". Capital in ", country, ", 1700-2010")) + theme(legend.position="bottom", legend.background = element_rect(fill="transparent")) + scale_x_continuous(breaks = seq(1700, 2020, by=20))
}
```

Now we make Figure F3.1:

```{r figF3_1, dependson="tabTS3.1", fig.width=8, fig.height=4}
run31_32(tab, num = "3.1", "Britain")
```

Figure F3.2:

```{r figF3_2, dependson="tabTS3.2", fig.width=8, fig.height=4}
run31_32(tab32, num = "3.2", "France")
```

In Figure 3.1 and 3.2 of Piketty, the axes are equally spaced, but correspond to different periods of time such as the first two tickmarks represent 1700-1750, a period of 50 years, whereas the last 2 tickmarks represent 1990-2010, spanning 30 years, and the ones before span 1970-1990, only 20 years.  

## Figures 3.3 and 3.4
Let's make another function to create the plot for 3.3 and 3.4
```{r func3_3,  fig.width=8, fig.height=4}
run_33_34 = function(tab, num, country){
fdat <- tab[, c("year","public_ass", "public_debt")]
names(fdat) <- c("year", "Public Assets", "Public Debt")
fdat <- melt(fdat, id.var="year")
names(fdat) <- c("year", "Capital", "percent")

qplot(year, percent, data=fdat, geom=c("line", "point"), colour = Capital, xlab= "Year", ylab= "Public assets and debt (% national income)") + scale_y_continuous(labels = percent, limits = c(0, 2.5)) + 
  ggtitle(paste0("Figure ", num, ". Public wealth in ", country, ", 1700-2010")) + theme(legend.position=c(0.5, .8), legend.background = element_rect(fill="transparent")) + scale_x_continuous(breaks = seq(1700, 2020, by=20))
}
```

Figure 3.3

```{r fig3_3, dependson="tabTS3.1", fig.width=8, fig.height=4}
run_33_34(tab, "3.3", "Britain")
```

Figure 3.4

```{r fig3_4, dependson="tabTS3.2", fig.width=8, fig.height=4}
run_33_34(tab32, "3.4", "France")
```


## Figures 3.5 and 3.6
Let's make another function to create the plot for 3.5 and 3.6
```{r func3_5, fig.width=8, fig.height=4}
run_35_36 = function(tab, num, country){
fdat <- tab[, c("year","nat", "public", "private")]
names(fdat) <- c("year", "National Capital (Public + Private)", "Public", "Private")
fdat <- melt(fdat, id.var="year")
names(fdat) <- c("year", "Capital_Type", "percent")

qplot(year, percent, data=fdat, geom=c("line", "point"), colour = Capital_Type, xlab= "Year", ylab= "National, private and public capital (% national income)") + scale_y_continuous(labels = percent, limits = c(-2, 9)) + 
  ggtitle(paste0("Figure ", num, ". Private and public capital in ", country, ", 1700-2010")) + theme(legend.position=c(0.5, .5), legend.background = element_rect(fill="transparent")) + scale_x_continuous(breaks = seq(1700, 2020, by=20))
}
```

Figure 3.5

```{r fig3_5, dependson="tabTS3.1", fig.width=8, fig.height=4}
run_35_36(tab, "3.5", "Britain")
```

Figure 3.6

```{r fig3_6, dependson="tabTS3.2", fig.width=8, fig.height=4}
run_35_36(tab32, "3.6", "France")
```