BHR hangs distribution and uptime correlation

BHR.R

# BHR hangs distribution and uptime correlation

library(ggplot2)
library(dplyr)

session <- read.csv("session.csv") %.% filter(uptime >= 0)

ggplot(aes(x=hangs), data=session) + 
  geom_histogram(aes(y=..density..), binwidth=1, colour="black", fill="blue") + 
  scale_x_continuous(limits=c(0, 50)) + #~ 90th quantile for sessions
  theme_bw() + labs(x="Number of hangs per session", y="Density", title="Hangs distribution")

ggplot(aes(x = uptime, y=hangs), data=filter(session, uptime < quantile(session$uptime, 0.99))[1:1000000,]) + 
  geom_point(alpha=0.01, position="jitter") + 
  stat_summary(fun.y = median, colour="red", geom="point") +
  scale_y_log10(breaks=c(10, 100, 1000, 10000, 100000, 1000000)) +
  scale_x_continuous() +
  labs(title="Median number of hangs vs uptime") +
  theme_bw()