R中的多线图

Question

I have a .csv file of the average life expectancy by country for the last 50 years. I am trying to create a graph of life expectancy by country, with the years 1960-2011 on the x axis, and the average life expectancy on the y axis. I only want to plot the top ten countries, each with their own line. I have researched every possible way to plot a multi line graph of the data I have and it seems to me that it is impossible with the way the data is formatted. My questions are:

1. Is it possible to create the desired graph with this data, given the way it is organized?
2. If the data has to be restructured, how should that be done? Is there a function in R to better organize data?

I was able to create the desired graph in Excel which is exactly what I'd like to do in R.

Here is a link to the lexp.csv file. https://drive.google.com/file/d/0BwsBIUlCf0Z3QVgtVGt4ampVcmM/view?usp=sharing

Answer 1

You are correct that the data would benefit from reorganization. This is a "wide to long" problem ie it would be better to have 3 columns: Country, Year and Age.

You can reformat the data using the tidyr package, process it using the dplyr package and plot using ggplot2 . So, assuming that you have read the CSV into R and have a data frame named lexp , you could try something like this:

library(dplyr)
library(tidyr)
library(ggplot2)

lexp %>% 
  # reformat from wide to long
  gather(Year, Age, -Country, convert = TRUE) %>%
  # select most recent year 
  filter(Year == max(Year)) %>%
  # sort by decreasing age 
  arrange(desc(Age)) %>% 
  # take the top 10 countries
  slice(1:10) %>% 
  select(Country) %>% 
  # join back to the original data
  inner_join(lexp) %>% 
  # reformat again from wide to long
  gather(Year, Age, -Country, convert = TRUE) %>% 
  # and plot the graph
  ggplot(aes(Year, Age)) + geom_line(aes(color = Country, group = Country)) +
    theme_dark() + theme(axis.text.x = element_text(angle = 90)) + 
    labs(title = "Life Expectancy") +
    scale_color_brewer(palette = "Set3")

Result:

Answer 2

library("reshape2")
library("ggplot2")

test_data_long <- melt(df, id="Country")  # convert to long format
testdata<-test_data_long[complete.cases(test_data_long),]
ggplot(data=testdata,
       aes(x=variable, y=value)) +
  geom_line(aes(color = Country, group = Country))