Homework week 3

library(plyr)
library(sqldf)
library(dplyr)
library(scales)
library(tcltk)
library(plotly)

pop.url <- file(paste(url,"population.csv", sep = ""), open="r" )
tb.url <- file(paste(url,"tb.csv", sep = ""),  open="r" ) 
population <- read.csv(pop.url, sep=",", header=TRUE, stringsAsFactors = FALSE)
tb_infections <- read.csv(tb.url, sep=",", header=FALSE, stringsAsFactors = FALSE)
close(pop.url, tb.url)
colnames(tb_infections) <- c("country","year","sex","child","adult","elderly")
tb_infections[tb_infections == "-1"] <- NA
tb_infections <- ddply(tb_infections,~country + year, summarise, adult = sum(adult), child = sum(child), elderly = sum(elderly))
tb_infections$count <- tb_infections$child + tb_infections$adult + tb_infections$elderly
data <- sqldf("SELECT DISTINCT p.country, p.year, t.count, p.population 
              FROM population p LEFT JOIN tb_infections t on t.country = p.country and t.year = p.year")
data$rate <- (data$count/data$population)
data <- data[,c(1,2,5)]
subset <- ddply(data, ~country, summarize, avgrt = mean(rate, na.rm = TRUE))
subset <- subset[order(-subset$avgrt),]
n <- 5
top_highrt <- head(subset$country, n = n)
data <- subset(data, data$country %in% top_highrt)
per_person <- 1000
title = sprintf("TB Rate per %s people for top %s countries", comma(per_person), n)
p <- plot_ly(data, x = year, y = (rate * per_person), color = country, type = "line",
showlegend = TRUE) %>% layout (title = title, yaxis = list(title = sprintf("Rate per %s people",per_person)), xaxis = list(title = "year"))
p