weekly_awesome_upgrades.R

library(data.table); library(dplyr); library(tidyr); library(ggplot2); library(xts)

## Import and Tidy
stripe <- read.table('~/Downloads/stripe.csv',sep=',',header=T)
stripe$start_date <- as.Date(as.POSIXlt(stripe$start, format="%Y-%m-%d %H:%M:%S"))
stripe$end_date <- as.Date(as.POSIXlt(stripe$ended_at, format="%Y-%m-%d %H:%M:%S"))

## Aggregate Awesome Start Dates and Count Users
awesome_start_date <- stripe %>%
  filter(plan_id == 'pro-annual' | plan_id == 'pro-monthly' | plan_id == 'awesome') %>%
  group_by(start_date) %>%
  summarise(users = n_distinct(id)) %>%
  mutate(cumulative = cumsum(users))
  
## Plot Daily Awesome Plan Starts
ggplot(filter(awesome_start_date,start_date > '2015-01-01'),aes(x=start_date,y=users)) +
  geom_point(size=1) +
  geom_line() +
  geom_smooth() +
  theme_minimal()
  
## Aggregate Daily Data to Weekly
aggregated_awesome <- as.xts(awesome_start_date$users,order.by=as.Date(awesome_start_date$start_date))
awesome_weekly <- apply.weekly(aggregated_awesome,sum)
awesome_df <- data.frame(date=index(awesome_weekly), coredata(awesome_weekly))
names(awesome_df) <- c('week','users')

## Plot Weekly Awesome Upgrades
ggplot(filter(awesome_df,week > '2015-01-01' & week < '2015-11-03'),aes(x=week,y=users)) +
  geom_point(size=3) +
  geom_line(size=1) +
  labs(x='Week',y='',title='Weekly Awesome Upgrades') +
  theme_minimal()
  
## Check Individual IDs of Awesome Upgrades in a Given Week
awesome_week_users <- stripe %>%
  filter((plan_id == 'pro-annual' | plan_id == 'pro-monthly') & (start_date >= '2015-10-01' & start_date <= '2015-10-07'))