project1_607.Rmd

---
title: "Project1_607"
author: "Alvaro Bueno"
date: "9/22/2017"
output: pdf_document
---
## build the data frames  
```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
results <- data.frame(id=integer(),name=character(),state=character(),points=double(),prescore=integer(), calcscore=double())
names(results) <- c('id','name','state', 'points', 'prescore', 'calcscore')
opponents = data.frame(id=integer(),opponents=character())

```

## extract the data from the file
```{r load-data, eval=TRUE}
library('stringr')
tmp <- url('https://github.com/delagroove/dataScience/raw/master/tournamentinfo.txt')
open(tmp, 'r')
count = 0
header <-readLines(tmp, n=4)
trim <- function (x) gsub("^\\s+|\\s+$", "", x)

while (length(data <- readLines(tmp, n=3)) > 0){
    count = count + 1
    #extract id
    id <- unlist(str_extract_all(data[1], "(?<= {3,4})[0-9]{1,2}(?= {1}| [a-zA-Z ]{6,})", FALSE))

    name <- unlist(str_extract_all(data[1], "(?<=[0-9]{1,2} {1}| )[a-zA-Z ]{6,}", FALSE))
    name <- trim(name)
    
    #extract states
    state <- unlist(str_extract_all(data[2], "[a-zA-Z]{2}(?= | [0-9]{8})", FALSE))
    
    #extract points
    points <- unlist(str_extract_all(data[1], "(?<=[0-9]{1,2} {1}| [a-zA-Z ]{6,32}|)[0-9].[0-9]", FALSE))
    
    #extract pre-score
    pres <- unlist(str_extract_all(data[2], "(?<=/ R: {1,2})[0-9]{3,4}(?= |P)", FALSE))
    df <- data.frame(id,name,state,points,pres, 0)
    names(df) <- c('id', 'name','state', 'points', 'prescore', 'calcscore')
    results <- rbind(results, df)
    
    #extract opponents
    opp <- unlist(str_extract_all(data[1], "(?<=[0-9]{1,2} {1}| [a-zA-Z ]{6,32}|[0-9].[0-9]  |(W|D|L) {2,3})[0-9]{1,}", FALSE))

    df2 <- data.frame(id, opp)
    names(df2) <- c('id','opponents')
    opponents <- rbind(opponents, df2)
    
}
```
## Fill out the calculated store
```{r}
for (id in 1:max(as.numeric(opponents$id))){
  sum <- 0
  for(opponent in opponents[opponents$id==id,'opponents']){
    pr <- results[results$id==opponent,'prescore']
    num_pr <- as.numeric(levels(pr))[pr]
    sum = sum + num_pr
  }
  
  calcscore = sum/length(opponents[opponents$id==id,'opponents'])
  results[results$id==id,'calcscore'] <- calcscore
}

close(tmp)
View(results)

#write results to CSV
write.csv(results, file = "results.csv")
```