Review_Datav4

## Clear environment
rm(list = ls())

## Load packages
library(dplyr)

## Use https://konklone.io/json/ to convert any individual jsonlines file into a CSV.
## All raw CSVs are available @
## https://drive.google.com/a/mst.edu/file/d/0B6wt0HET6pITVHJSZFo3YTNMWWc/view?usp=sharing

## Read CSV into R
ARMA3_DF <- read.csv("G:\\Users\\Chase\\Desktop\\IST3420Project\\Data\\ARMA_3.csv", header = TRUE, sep = ",")
CS_DF <- read.csv("G:\\Users\\Chase\\Desktop\\IST3420Project\\Data\\Counter_Strike.csv", header = TRUE, sep = ",")
CSGO_DF <- read.csv("G:\\Users\\Chase\\Desktop\\IST3420Project\\Data\\Counter_Strike_GO.csv", header = TRUE, sep = ",")
DOTA2_DF <- read.csv("G:\\Users\\Chase\\Desktop\\IST3420Project\\Data\\Dota_2.csv", header = TRUE, sep = ",")
TF2_DF <- read.csv("G:\\Users\\Chase\\Desktop\\IST3420Project\\Data\\TF2.csv", header = TRUE, sep = ",")
WarF_DF <- read.csv("G:\\Users\\Chase\\Desktop\\IST3420Project\\Data\\Warframe.csv", header = TRUE, sep = ",")

## Add "paid/free" column to each individual dataset
## This will help us compare later once the datasets are joined together.

ARMA3_DF$Monetization <- c("Paid")
CS_DF$Monetization <- c("Paid")
CSGO_DF$Monetization <- c("Paid")
DOTA2_DF$Monetization <- c("Free")
TF2_DF$Monetization <- c("Free")
WarF_DF$Monetization <- c("Paid")

## Add "game" column to each individual dataset
## This will help us compare later once the datasets are joined together.

ARMA3_DF$Game_Name <- c("ARMA 3")
CS_DF$Game_Name <- c("Counterstrike")
CSGO_DF$Game_Name <- c("Counterstrike: Global Offensive")
DOTA2_DF$Game_Name <- c("DOTA 2")
TF2_DF$Game_Name <- c("Team Fortress 2")
WarF_DF$Game_Name <- c("Warframe")

## Add "genre" column to each individual dataset
## This will help us compare later once the datasets are joined together.

ARMA3_DF$Genre <- c("Shooter")
CS_DF$Genre <- c("Shooter")
CSGO_DF$Genre <- c("Shooter")
DOTA2_DF$Genre <- c("MOBA")
TF2_DF$Genre <- c("Shooter")
WarF_DF$Genre <- c("MMORPG")

## Time to merge the dataframes

Review_Data <- rbind(ARMA3_DF, CS_DF, CSGO_DF, DOTA2_DF, TF2_DF, WarF_DF)

## We now remove columns that aren't necessary for our functions

Review_Data <- Review_Data %>% select(-c(num_groups,orig_url,num_badges,review_url,num_found_funny,date_updated,
                                         num_workshop_items,date_posted,found_helpful_percentage,num_voted_helpfulness,
                                         achievement_progress.num_achievements_percentage,achievement_progress.num_achievements_attained,
                                         achievement_progress.num_achievements_possible,profile_url,num_found_helpful,steam_id_number,
                                         friend_player_level,num_found_unhelpful,username,num_guides,num_comments,num_reviews,num_games_owned,num_screenshots))

## Rename columns for clarity

Review_Data <- Review_Data %>% rename(Recent_Played_Time = total_game_hours_last_two_weeks,Review = review
                                      ,Total_Played_Time = total_game_hours,Rating = rating,Friends = num_friends)

## Replace NULL values with 0s for clarity/comparison

Review_Data <- Review_Data %>% mutate(Friends = ifelse(is.na(Friends),0,Friends))

## Remaining factors are converted to characters

Review_Data <- rapply(Review_Data, as.character, classes="factor", how="replace")

## Add new column "Review_Char_Count" to track how many characters were in a review

Review_Data$Review_Char_Count <- c(nchar(Review_Data$Review, allowNA = TRUE))

## Export data as CSV

write.csv(Review_Data, file = "ReviewData_v4.csv")