Analysis of StackOverflow activity trends

gistfile1.txt

library(dplyr)
library(ggplot2)
library(reshape2)

# used the query here:
# http://data.stackexchange.com/stackoverflow/query/213011/metrics-of-question-activity-over-time

dat = read.csv("permonth_data.csv")

dat = dat %>% mutate(Date=as.Date(paste(Year, Month, "15", sep="-")))
# remove the first few months (very different site then) and the last month (since the data
# may be slightly out of date)
dat = dat %>% filter(Year > 2008) %>% filter(Year < 2014 | Month < 7)

print(ggplot(dat, aes(Date, Questions)) + geom_line() + ggtitle("Questions Per Month"))

dat.m = melt(dat, id=c("Year", "Month", "Date", "Questions"))

dat.m = dat.m %>% mutate(normalized=value / Questions)

ggsave("Questions.png")

print(ggplot(dat.m, aes(Date, normalized)) + geom_line() + facet_wrap(~ variable, scale="free_y") +
          ggtitle("Average Per Question") + ylab("Metric per question"))

ggsave("Averages.png")

print(ggplot(dat.m, aes(Date, value)) + geom_line() + facet_wrap(~ variable, scale="free_y") +
          ggtitle("Total (Unnormalized)") + ylab("Number"))

ggsave("Totals.png")