Wickets

library(ggplot2)
library(dplyr)
library(plotly)
library(knitr)
library(rjson)
library(tidyverse)
library(readr)
library(data.table)

FILE_PATH = params$FILE_PATH
FILTER_YEAR = params$FILTER_YEAR
PROCESS_DATA = params$PROCESS_DATA
PROCESS_LINK = params$PROCESS_LINK
PROCESS_OUTPUT = params$PROCESS_OUTPUT

readmePath <- paste(FILE_PATH, "/", "README.txt", sep = "")
readmeData <- read_lines(readmePath, skip = 24)

matches <- tribble( ~ Date, ~ Id)
for (d in readmeData) {
  items = strsplit(d, " - ")[[1]]
  year <- str_split(items[[1]], "-")[[1]][[1]]
  matches <- matches %>% add_row(Date = year, Id = items[[5]])
}

if (FILTER_YEAR != "NONE") {
  filteredMatches <- matches %>% filter(Date == FILTER_YEAR)
} else {
  filteredMatches <- matches
}
files <-
  filteredMatches %>% mutate(path = paste(FILE_PATH, "/", Id, ".json", sep = ""))

nameChanges <- tribble(
  ~ Old, ~ New,
  "Delhi Daredevils", "Delhi Capitals",
  "Rising Pune Supergiant", "Rising Pune Supergiants",
  "Kings XI Punjab", "Punjab Kings"
)

processName <- function (oldName) {
  if (oldName %in% nameChanges$Old) {
    i <- which(oldName == nameChanges$Old)
    nameChanges$New[[i]]
  } else {
    oldName
  }
}

if (PROCESS_DATA) {
  matchData <- tribble(~ FilePath, ~ RunsPerWicket, ~ Year, ~ Team, ~ Inning)
  for (file in files$path) {
    result <- fromJSON(file = file, simplify = TRUE)
    date <- strsplit(result$info$dates[[1]], "-")[[1]][[1]]
    inningNumber <- 1
    for (inning in result$innings) {
      wickets <- c()
      runs <- 0
      prevRuns <- 0
      for (over in inning$overs) {
        for (delivery in over$deliveries) {
          if (length(delivery$wickets) > 0) {
            wickets <- append(wickets, runs - prevRuns)
            prevRuns <- runs
          } else {
            runs <- runs + delivery$runs$total
          }
        }
      }
      matchData <- add_row(matchData, FilePath = file, RunsPerWicket = list(wickets), Year = date, Team = processName(inning$team), Inning = inningNumber)
      inningNumber <- inningNumber + 1
    }
  }
} else {
  initData <- as_tibble(fread(PROCESS_LINK)) %>% mutate(R = RunsPerWicket)
  matchData <- initData %>% 
    mutate(RunsPerWicket = lapply(strsplit(R, split = "\\|"), as.numeric))
}

cleaned <- na.omit(matchData) %>% 
  mutate(SumRuns = sapply(RunsPerWicket, sum)) %>%
  filter(SumRuns > 0)

if (PROCESS_OUTPUT && PROCESS_DATA) {
  fwrite(cleaned, PROCESS_LINK)
}

tData <- tribble(~ Wicket, ~ Value, ~ Year, ~ PercentValue)
for (y in unique(cleaned$Year)) {
  sumP <- 0
  sumL <- 0
  yearFiltered <- cleaned %>% filter(as.numeric(Year) == as.numeric(y))
  yearSum <- mean(yearFiltered$SumRuns)
  ys <- sum(yearFiltered$SumRuns)
  for (i in 1:10) {
    r <- c()
    totals <- c()
    for (ia in 1:length(yearFiltered$RunsPerWicket)) {
      run <- yearFiltered$RunsPerWicket[[ia]]
      if (length(run) >= i) {
        r <- append(r, run[[i]])
        totals <- append(totals, yearFiltered$SumRuns[[i]])
      }
    }
    sumP <- sumP + (mean(r)/yearSum)
    sumL <- sumL + (sum(r)/ys)
    tData <- tData %>% add_row(Wicket = i, Value = mean(r, na.rm = TRUE), Year = y, PercentValue = (sum(r)/ys))
  }
}

teamData <- tribble(~ Wicket, ~ Value, ~ PercentValue, ~ Team)
for (y in unique(cleaned$Team)) {
  sumP <- 0
  sumL <- 0
  teamF <- cleaned %>% filter(Team == y)
  yearSum <- mean(teamF$SumRuns)
  ys <- sum(teamF$SumRuns)
  for (i in 1:10) {
    r <- c()
    totals <- c()
    for (ia in 1:length(teamF$RunsPerWicket)) {
      run <- teamF$RunsPerWicket[[ia]]
      if (length(run) >= i) {
        r <- append(r, run[[i]])
        totals <- append(totals, teamF$SumRuns[[i]])
      }
    }
    sumP <- sumP + (mean(r)/yearSum)
    sumL <- sumL + (sum(r)/ys)
    teamData <- teamData %>% add_row(Wicket = i, Value = mean(r, na.rm = TRUE), Team = y, PercentValue = (sum(r)/ys))
  }
}

# Embed 1
summary(cleaned %>% select(FilePath, Year, Inning, SumRuns))

# Embed 2
meanData <- tData %>% 
  group_by(Wicket) %>% 
  summarise(Wicket = mean(Wicket), Value = mean(Value), PercentValue = mean(PercentValue))
pmean <- ggplot(meanData, aes(x = Wicket, y = PercentValue)) +
  geom_line() +
  geom_point() +
  xlab("Wicket") +
  ylab("% of Total Runs")
ggplotly(pmean)

# Embed 3
runsLengths <- matchData %>% filter(Inning < 3) %>% rowwise() %>% mutate(NumberWickets = length(RunsPerWicket))
prl <- ggplot(runsLengths, aes(x = NumberWickets)) +
  geom_histogram(bins = 10)
ggplotly(prl)

# Embed 4
pteams <- ggplot(teamData, aes(x = Wicket, y = PercentValue, colour = Team)) +
  geom_line() +
  geom_point() +
  xlab("Wicket") +
  ylab("% of Total Runs")
ggplotly(pteams)

# Embed 5
ptimeC <- ggplot(tData, aes(x = Wicket, y = PercentValue, colour = factor(Year))) +
  geom_line() +
  geom_point() +
  xlab("Wicket") +
  ylab("% of Total Runs")
ggplotly(ptimeC)

compileYears <- function (years) {
  a <- tData %>% 
    filter(Year %in% years) %>% 
    group_by(Wicket) %>% 
    summarise(
      AvgVal = mean(PercentValue), 
      YearVal = paste(years[[1]], "to", years[[length(years)]], sep = " ")
    )
  a
}

combinedPlot <- function (year1, year2) {
  comp1 <- compileYears(year1)
  comp2 <- compileYears(year2)
  
  comData <- bind_rows(comp1, comp2)

  pcom <- ggplot(comData, aes(x = Wicket, y = AvgVal, colour = YearVal)) +
    geom_line() +
    geom_point() +
    xlab("Wicket") +
    ylab("% of Total Runs")
  ggplotly(pcom)
}

# Embed 6
combinedPlot(c(2008, 2010, 2011, 2012, 2013, 2014), c(2015, 2016, 2017, 2018, 2019))

# Embed 7
pteams <- ggplot(teamData, aes(x = Wicket, y = PercentValue, colour = Team)) +
  geom_line() +
  geom_point() +
  xlab("Wicket") +
  ylab("% of Total Runs")
ggplotly(pteams)

# Embed 8
order1 <- teamData %>% 
  filter(Wicket == 1 | Wicket == 2 | Wicket == 3) %>%
  group_by(Team) %>%
  summarise(FirstThree = sum(PercentValue) * 100)

order2 <- teamData %>% 
  filter(Wicket == 4 | Wicket == 5 | Wicket == 6) %>%
  group_by(Team) %>%
  summarise(MidThree = sum(PercentValue) * 100) %>%
  select(MidThree)

order3 <- teamData %>% 
  filter(Wicket == 7 | Wicket == 8 | Wicket == 9) %>%
  group_by(Team) %>%
  summarise(LastFour = sum(PercentValue) * 100) %>%
  select(LastFour)

combined <- bind_cols(order1, order2, order3) %>% arrange(desc(FirstThree))

kable(combined)

#END