Homework #3: Data Aggregation

Setup the data and libraries

library(plyr)
library(xtable)
# Import the Gapminder dataset
gDat <- read.delim("gapminderDataFiveYear.txt")

Part 1: Presenting minimum and maximum GDP per capita in “wide” format

table1 <- ddply(gDat, ~continent, summarize, minimum.GDP.per.capita = min(gdpPercap), 
    `maximum GDP per capita` = max(gdpPercap))
table1 <- arrange(table1, minimum.GDP.per.capita)
print(xtable(table1), type = "html", include.rownames = F)

Results:

Africa has the country lowest minimum and maximum GDP.

Part 2: Spread of GDP per capita in different continents

table2 <- ddply(gDat, ~continent, summarize, GDP.standard.deviation = sd(gdpPercap), 
    GDP.median.absolute.deviation = mad(gdpPercap), GDP.interquartile.range = IQR(gdpPercap))
table2 <- arrange(table2, GDP.standard.deviation)
print(xtable(table2), type = "html", include.rownames = F)

Results:

Asia has the highest standard devaition of GDP, while Europe has the highest variance for median absolute deviation and interquartile range. This is likely due to outliers in Asia.

Part 3: How life expectancy changes over years

table3 <- ddply(gDat, ~year, summarize, mean.life.expectancy = mean(lifeExp), 
    mean.life.expectancy.trimmed.20.percent = mean(lifeExp, trim = 0.1), mean.life.expectancy.trimmed.40.percent = mean(lifeExp, 
        trim = 0.2))
print(xtable(table3), type = "html", include.rownames = F)

Results:

Mean life expectancy increases over time. In the 50's, trimming outliers decreases life expectancy, while in the 60's onward, the reverse is true. In the 50's the mean life expectancy was driven upward by high outliers. Past that, low outliers reduced the mean life expectancy.