Introduction

This analysis is the result of an excercise made in Udacity’s Course: Exploratory Data Analysis with R

The goal is to analyze the Forest Coverage dataset to study the increment in its proportion. The goal is to answer questions like:
- Which countries have more Forest Coverage?
- Which countries cares the most about Preserving its Forest Coverage?

Excercise Instructions: Gapminder Data

The Gapminder website contains over 500 data sets with information about the world’s population. Your task is to download a data set of your choice and create 2-5 plots that make use of the techniques from Lesson 3.

You might use a simple histogram, a boxplot split over a categorical variable, or a frequency polygon. The choice is yours!

Link to the Gapminder website.

Once you’ve completed your investigation, create a post in the discussions that includes:
1. any questions you answered, your observations, and summary statistics
2. snippets of code that created the plots
3. links to the images of your plots

Data Analysis

Load necesary Libraries

library(reshape2)
library(plyr)
library(ggplot2)
library(rworldmap)
library(RColorBrewer)

Download and Cleaning data

setwd("~/Repos/data-analysis-with-r/")
file.path <- "./data/indicator_forest coverage.csv"
data <- read.csv(file.path,col.names = c("Country","1990","2000","2005"))

data <- melt(data, id.vars = c("Country"))

## Warning: attributes are not identical across measure variables; they will
## be dropped

names(data) <- c("Country","Year","Forest.Coverage")

data$Year <- gsub("X","",data$Year)
data$Year <- as.factor(data$Year)

data$Forest.Coverage <- gsub(",",".",data$Forest.Coverage)
data$Forest.Coverage <- as.numeric(data$Forest.Coverage,na.action)

data <- data[complete.cases(data),]

Data Processing

ggplot(data) + 
  geom_boxplot(aes(x = Year, y = Forest.Coverage, fill = Year), alpha = 0.5) + 
  scale_y_continuous(breaks = seq(0,60,5)) +
  coord_cartesian(ylim = c(10,60)) + 
  ylab("Average Forest Coverage (%)") + 
  ggtitle("Average Forest Coverage (%) by Year")

We can deduce that mean of the percent of Forest Coverage accross all the Countries decreased from ~30% to ~28% between 1990 and 2005.

Argentina:

argentina <- subset(data, Country == "Argentina")
argentina

##       Country Year Forest.Coverage
## 215 Argentina 1990           12.88
## 444 Argentina 2000           12.34
## 673 Argentina 2005           12.07

ggplot(argentina) + 
  geom_line(aes(x=Year,y=Forest.Coverage, group = Country), stat= "identity",
            size = 1, color = "red") +
  geom_boxplot(data = data , aes(x=Year,y=Forest.Coverage, fill = Year), alpha = 0.25) + 
  ggtitle("Forest Coverage (%) Changes in Argentina between 1990 and 2005\n 
          compared with the average of the World") + 
  ylab("Average Forest Coverage (%)") 

Argentina’s Forest Coverage went down from 12.88% to 12.07%. Argentina’s values are limiting the First Quantile of the entire population.

South America

south.america <- c("Argentina", "Chile", "Uruguay", "Paraguay", "Brazil",
                   "Peru", "Ecuador", "Bolivia", "Venezuela", "Colombia",
                   "Suriname", "French Guiana", "Guyana")
south.america.data <- subset(data, Country %in% south.america)

ggplot(south.america.data) + 
  geom_line(aes(x=Year,y=Forest.Coverage, group = Country, 
                color = Country),
            stat= "identity", size = 1) +
  ggtitle("Forest Coverage (%) in South America Countries\n 
           compared with the average of the region.") + 
  geom_boxplot(aes(x=Year,y=Forest.Coverage, fill = Year), alpha = 0.25) + 
  ylab("Forest Coverage (%)") +
  scale_y_continuous(breaks = seq(0,70,5))

Some conclusions:
- There are a lot of differences between the top countries, Brazil and Colombia, and Uruguay and Chile. Almost a 50% of difference.
- Only 2 countries has increased it’s Forest Coverage, Uruguay and Chile.
- The mean of Forest Coverage in South America decreased from ~58% to ~50%. This difference is bigger than the average of the World.
- Countries like Brazil, Colombia and Venezuela has almost a 60% of Forest Coverage!

G20 Countries

I’ve excluded European Union and South Korea due to missing values in the Data Set

g20 <- c("Argentina","Australia", "Brazil","Canada","China","France",
                   "Germany", "India", "Indonesia","Italy", "Japan","Mexico",
                   "Russia", "Saudi Arabia", "South Africa",
                   "Turkey", "United Kingdom", "United States")
g20.data <- subset(data, Country %in% g20)

ggplot(g20.data) + 
  ggtitle("Forest Coverage (%) Changes in G20 Countries.\n 
           between 1990 and 2005.") + 
  geom_boxplot(aes(x=Year,y=Forest.Coverage, fill = Year), alpha = 0.25) + 
  ylab("Forest Coverage (%)") +
  scale_y_continuous(breaks = seq(0,70,5))

We can deduce that mean of the Forest Coverage accross all the g20 Countries increased from ~27.5% to ~30% between 1990 and 2005. Since the World tendency is to decrease, Which Countries in the G20 Group have increased it’s Forest Coverage?

G20 Countries that increased its Forest Coverage

Now I want to calculate the variance in the Forest Coverage in the G20 Countries to determine which ones increased and which ones decreased.

# Calculate de Difference Between FC in 2005 and FC in 1990.
Changes <- subset(g20.data, Year == 2005)$Forest.Coverage - subset(g20.data, Year == 1990)$Forest.Coverage
# Get the list of Countries
Country <- subset(g20.data, Year == 2005)$Country
# Create the Data Frame
g20.Changes <- data.frame(Country, Changes)
# Order the Data Frame by Changes
g20.Changes <- arrange(g20.Changes,Changes,decreasing = TRUE)
# Reset the Factors levels so ggplot can plot in correct order
g20.Changes$Country <- factor(g20.Changes$Country,
                                      levels = g20.Changes$Country) 
g20.Changes

##           Country Changes
## 1           Italy    5.43
## 2           China    4.30
## 3          France    1.84
## 4           India    1.26
## 5  United Kingdom    0.97
## 6         Germany    0.96
## 7          Turkey    0.64
## 8   United States    0.48
## 9    South Africa    0.00
## 10   Saudi Arabia    0.00
## 11         Canada    0.00
## 12         Russia   -0.01
## 13          Japan   -0.23
## 14      Australia   -0.55
## 15      Argentina   -0.81
## 16         Mexico   -2.50
## 17         Brazil   -5.00
## 18      Indonesia  -15.50

ggplot(g20.Changes) + 
  geom_bar(aes(x=Country,y=Changes), fill = "blue4", stat = "identity") +
  scale_y_continuous(breaks = seq(-20,15,1)) + 
  ggtitle("Changes in Forest Coverage (%)\n in G20 countries from 1990 to 2005") +
  ylab("Changes in Forest Coverage (%)") +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

## Warning: Stacking not well defined when ymin != 0

Some interesting results:
- Italy, China, France, India, United Kingdom, Germany, Turkey and United States increased their Forest Coverage from 1990 to 2005
- Japan, Australia, Argentina, Mexico, Brazil, and Indonesia decreased their Forest Coverage between 1900 and 2005
- Wondering about Indonesia? Deforestation in Indonesia

Top Countries in the World 2005

Which are the Countries in the World with most Forest Coverage?

top.countries.2005 <- subset(data, Year == 2005)
top.countries.2005 <- top.countries.2005[order(top.countries.2005$Forest.Coverage, 
                                               decreasing = TRUE),]

top.countries.2005$Country <- factor(top.countries.2005$Country, 
                                     levels = top.countries.2005$Country)
# Plot Top 20
ggplot(top.countries.2005[1:20,]) + 
  geom_bar(aes(x=Country, y = Forest.Coverage), fill = "blue4", stat = "identity") +
  ylab("Forest Coverage in 2005") + 
  ggtitle("20 Countries with more Forest Coverage (%) in the World in 2005.") +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

• Learn a little bit more about Pitcairng
  
• Some info about Suriname

Which are the Countries in the World with less Forest Coverage?

# Plot Top 20
ggplot(tail(top.countries.2005, 20)) + 
  geom_bar(aes(x=Country, y = Forest.Coverage), fill = "blue4", stat = "identity") +
  ylab("Forest Coverage in 2005") + 
  ggtitle("20 Countries with less Forest Coverage (%) in the World in 2005.") +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

Top/Bottom Changes in Forest Coverage in the World

Which countries has increased/decreased its Forest Coverage between 1990 and 2005?

Country <- as.character(unique(data$Country))
Changes <- subset(data, Year == 2005)$Forest.Coverage - subset(data, Year == 1990)$Forest.Coverage

countries.Changes <- data.frame(Country, Changes)
countries.Changes <- arrange(countries.Changes,Changes,decreasing = T)
countries.Changes$Country <- factor(countries.Changes$Country,
                                      levels = countries.Changes$Country) 

ggplot(countries.Changes[1:20,]) + 
  geom_bar(aes(x=Country,y=Changes), fill = "blue4", stat = "identity") +
  scale_y_continuous(breaks = seq(-11,15,1)) + 
  ggtitle("Top 20 Changes in Forest Coverage (%)\n from 1990 to 2005") +
  ylab("Changes in Forest Coverage (%)") +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

ggplot(tail(countries.Changes, 20)) + 
  geom_bar(aes(x=Country,y=Changes), fill = "blue4", stat = "identity") +
  scale_y_continuous(breaks = seq(-25,15,1)) + 
  ggtitle("Bottom 20 Changes in Forest Coverage (%)\n from 1990 to 2005") +
  ylab("Changes in Forest Coverage (%)") +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

## Warning: Stacking not well defined when ymin != 0

top.bottom <- rbind(countries.Changes[1:10,], tail(countries.Changes, 10))
ggplot(top.bottom) + 
  geom_bar(aes(x=Country,y=Changes), fill = "blue4", stat = "identity") +
  scale_y_continuous(breaks = seq(-25,15,1)) + 
  ggtitle("Top 10 and Bottom 10 changes in Forest Coverage (%)\n from 1990 to 2005") +
  ylab("Changes in Forest Coverage (%)") +
  theme(axis.text.x = element_text(angle = 45, hjust = 1))

## Warning: Stacking not well defined when ymin != 0

• Wow! Honduras has reduced its Forest Coverage by almost a 25%!

Fancy World Map Graph

Of course there should be a Fancy World Map with some info. The map show the changes in percent of Forest Coverage between 1990 and 2005

colourPalette <- brewer.pal(7,"RdYlGn")

sPDF <- joinCountryData2Map( countries.Changes, joinCode = "NAME", 
                             nameJoinColumn = "Country")

## 201 codes from your data successfully matched countries in the map
## 13 codes from your data failed to match with a country code in the map
## 43 codes from the map weren't represented in your data

mapParams <- mapCountryData( sPDF, nameColumnToPlot="Changes",
                colourPalette = colourPalette,
                mapTitle = "World changes in Forest Coverage (%),
                per Country between 1990 and 2005",
                oceanCol = "lightblue",
                missingCountryCol = "white",
                addLegend = FALSE)
do.call( addMapLegend
 , c(mapParams
 , legendLabels="all"
 , legendWidth=0.5 ))