# install.packages("data.table")
library("data.table")

outcome <- data.table::fread('outcome-of-care-measures.csv')
outcome[, (11) := lapply(.SD, as.numeric), .SDcols = (11)]
outcome[, lapply(.SD
                 , hist
                 , xlab= "Deaths"
                 , main = "Hospital 30-Day Death (Mortality) Rates from Heart Attack"
                 , col="lightblue")
        , .SDcols = (11)]

best <- function(state, outcome) {
  
  out_dt <- data.table::fread('outcome-of-care-measures.csv')
  
  outcome <- tolower(outcome)
   
  chosen_state <- state 

  if (!chosen_state %in% unique(out_dt[["State"]])) {
    stop('invalid state')
  }
  
  if (!outcome %in% c("heart attack", "heart failure", "pneumonia")) {
    stop('invalid outcome')
  }
  
  setnames(out_dt
           , tolower(sapply(colnames(out_dt), gsub, pattern = "^Hospital 30-Day Death \\(Mortality\\) Rates from ", replacement = "" ))
  )
  
  out_dt <- out_dt[state == chosen_state]
  
  col_indices <- grep(paste0("hospital name|state|^",outcome), colnames(out_dt))
   
  out_dt <- out_dt[, .SD ,.SDcols = col_indices]
  
  out_dt[, outcome] <- out_dt[,  as.numeric(get(outcome))]
  
  out_dt <- out_dt[complete.cases(out_dt),]
   
  out_dt <- out_dt[order(get(outcome), `hospital name`)]
  
  return(out_dt[, "hospital name"][1])

}

rankhospital <- function(state, outcome, num) {
    
    data <- read.csv("outcome-of-care-measures.csv", colClasses = "character")
    data <- data[c(2, 7, 11, 17, 23)]
    names(data)[1] <- "name"
    names(data)[2] <- "state"
    names(data)[3] <- "heart attack"
    names(data)[4] <- "heart failure"
    names(data)[5] <- "pneumonia"

    outcomes = c("heart attack", "heart failure", "pneumonia")
    if( outcome %in% outcomes == FALSE ) {
    	stop("invalid outcome")
    }

    states <- data[, 2]
    states <- unique(states)
    if( state %in% states == FALSE ) {
    	stop("invalid state")
    }
    
    if( num != "best" && num != "worst" && num%%1 != 0 ) {
    	stop("invalid num")
    }
    
    data <- data[data$state==state & data[outcome] != 'Not Available', ]

    data[outcome] <- as.data.frame(sapply(data[outcome], as.numeric))
    data <- data[order(data$name, decreasing = FALSE), ]
    data <- data[order(data[outcome], decreasing = FALSE), ]

    vals <- data[, outcome]
    if( num == "best" ) {
        rowNum <- which.min(vals)
    } else if( num == "worst" ) {
        rowNum <- which.max(vals)
    } else {
        rowNum <- num
    }

    data[rowNum, ]$name
}

rankall <- function(outcome, num = "best") {
    
    data <- read.csv("outcome-of-care-measures.csv", colClasses = "character")
    data <- data[c(2, 7, 11, 17, 23)]
    names(data)[1] <- "name"
    names(data)[2] <- "state"
    names(data)[3] <- "heart attack"
    names(data)[4] <- "heart failure"
    names(data)[5] <- "pneumonia"

    
    outcomes = c("heart attack", "heart failure", "pneumonia")
    if( outcome %in% outcomes == FALSE ) stop("invalid outcome")

    if( num != "best" && num != "worst" && num%%1 != 0 ) stop("invalid num")

    data <- data[data[outcome] != 'Not Available', ]
    
    data[outcome] <- as.data.frame(sapply(data[outcome], as.numeric))
    data <- data[order(data$name, decreasing = FALSE), ]
    data <- data[order(data[outcome], decreasing = FALSE), ]

    getHospByRank <- function(df, s, n) {
        df <- df[df$state==s, ]
        vals <- df[, outcome]
        if( n == "best" ) {
            rowNum <- which.min(vals)
        } else if( n == "worst" ) {
            rowNum <- which.max(vals)
        } else {
            rowNum <- n
        }
        df[rowNum, ]$name
    }
    
    states <- data[, 2]
    states <- unique(states)
    newdata <- data.frame("hospital"=character(), "state"=character())
    for(st in states) {
        hosp <- getHospByRank(data, st, num)
        newdata <- rbind(newdata, data.frame(hospital=hosp, state=st))
    }

    newdata <- newdata[order(newdata['state'], decreasing = FALSE), ]
    newdata
}