bayesball

##################################################
# collect all pitches of Cliff Lee for 2013 season
# using pitchRx package
###################################################

# create database of 2013 pitches

# library(dplyr)
# my_db <- src_sqlite("pitchRx.2013", create = TRUE)
# library(pitchRx)

bayesball

# Cliff Lee's pitches in first game of 2014 season
# March 31, 2014

# scrape all Gameday data for that day

library(pitchRx)
d <- scrape("2014-03-31", "2014-03-31")

# pick out the Cliff Lee pitch data
library(dplyr)

bayesball

##########################################
# the first pitch effect
# preview of Chapter 7 material in ABDWR
##########################################

# load in Retrosheet 2013 play-by-play data
# (have to download Retrosheet data and add run values)

load("pbp2013.Rdata")

bayesball

# 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="")

bayesball

#----------------------------------------------
# R code for Bayesian Computation Webinar
# Jim Albert - June 12, 2014
# albert@bgsu.edu
#----------------------------------------------

#----------------------------------------------
# PART I:  Bayesian Robust Regression Modeling
# Require packages foreign, LearnBayes, coda, ggplot2
#----------------------------------------------

bayesball

#----------------------------------------------
# R code for Bayesian Computation Webinar
# Jim Albert - June 12, 2014
# albert@bgsu.edu
#----------------------------------------------

#----------------------------------------------
# PART II:  Posterior-Predictive Model Checking
# Require packages pscl, arm, MASS
#----------------------------------------------

bayesball

#----------------------------------------------
# R code for Bayesian Computation Webinar
# Jim Albert - June 12, 2014
# albert@bgsu.edu
#----------------------------------------------
 
#---------------------------------------------------------------------
# PART III:  INTRODUCTION TO JAGS
# Negative binomial regression
# Example from Jackman, 2009

bayesball

#----------------------------------------------
# R code for Bayesian Computation Webinar
# Jim Albert - June 12, 2014
# albert@bgsu.edu
#----------------------------------------------

##################################################
# PART IV:  JAGS for Multilevel Modeling
# Require R packages Lahman, rjags, ggplot2
##################################################

bayesball

# read in variable names for game log dataset

Headers <- c("Date",                       "DoubleHeader",              
"DayOfWeek",                  "VisitingTeam",              
"VisitingTeamLeague",         "VisitingTeamGameNumber",    
"HomeTeam",                   "HomeTeamLeague"         ,   
"HomeTeamGameNumber",         "VisitorRunsScored"       ,  
"HomeRunsScore",              "LengthInOuts",              
"DayNight",                   "CompletionInfo",            
"ForfeitInfo",                "ProtestInfo"    ,           

bayesball

library(Lahman)
library(dplyr)

fielding.2b <- filter(Fielding, POS=="2B", yearID >= 1960)

rf.season.2b <- summarize(group_by(fielding.2b, yearID),
              RF.9 = 9 * (sum(PO, na.rm=TRUE) + 
                          sum(A, na.rm=TRUE)) / 
                    (sum(InnOuts / 3, na.rm=TRUE)))