Graphs of Cubs and White Sox Median Attendance for Recent Seasons

cubs.attendance.R

# Cubs / White Sox attendance study
# assume that Retrosheet game log files 
# gl1900.txt, gl1901.txt, ... are
# contained in folder gamelogs
# the files gl1954.txt through gl2013.txt are used in this code

get.data <- function(team, Seasons){
  data <- NULL
  for (year in Seasons){
     filename <- paste("gamelogs/gl",year,".txt", sep="")
     d <- read.csv(filename, header=FALSE)
     headers <- read.csv("gamelogs/game_log_header.csv")
     names(d) <- names(headers)
     data <- rbind(data, subset(d, HomeTeam==team))}
  data$Season <- substr(data$Date, 1, 4)
  require(dplyr)
  S <- summarize(group_by(data, Season),
                 Median=median(Attendence))
  S$Season <- as.numeric(S$Season)
  S
}

cubs <- get.data("CHN", 1964:2013)
sox <- get.data("CHA", 1964:2013)

library(ggplot2)
ggplot(cubs, aes(Season, Median)) +
  geom_point(size=3, color="red") + 
  geom_smooth(method="loess", span=0.25, size=2) +
  labs(title="Cubs Median Attendance:  1954 - 2013") +
  theme(plot.title = element_text(size = rel(2))) +
  theme(axis.title = element_text(size = rel(2))) +
  theme(axis.text = element_text(size = rel(2))) + 
  annotate("rect", xmin=1988, xmax=2013, ymin=0, ymax=45000,
           alpha=0.2, fill="yellow")

ggplot(sox, aes(Season, Median)) +
  geom_point(size=3, color="red") + 
  geom_smooth(method="loess", span=0.25, size=2) +
  labs(title="White Sox Median Attendance:  1954 - 2013") +
  theme(plot.title = element_text(size = rel(2))) +
  theme(axis.title = element_text(size = rel(2))) +
  theme(axis.text = element_text(size = rel(2)))

# look at ratio median Cubs attendance divided by 
# median Sox attendance

bothteams <- merge(cubs, sox, by="Season")
bothteams$Ratio <- with(bothteams, Median.x / Median.y)

ggplot(bothteams, aes(Season, Ratio)) +
  geom_point(size=3, color="red") + 
  geom_smooth(method="loess", span=0.25, size=2) +
  labs(title="Ratio of Cubs / Sox Attendance:  1954 - 2013") +
  theme(plot.title = element_text(size = rel(2))) +
  theme(axis.title = element_text(size = rel(2))) +
  theme(axis.text = element_text(size = rel(2))) +
  geom_hline(yintercept=1)  +
  ylim(-.5, 4.5)