RunningShiny.R

setwd("/Users/matthewfisher/Dropbox/Experiments/Tracking Method/")
install.packages('rsconnect')
rsconnect::setAccountInfo(name='mfisher',
                          token='1715DA04549BEF565B303A8E19037AE7',
                          secret='hh3j7rycFcoUspfWsv2f9vgepwAY/EfTYwA+8eS/')

rsconnect::deployApp('/Users/matthewfisher/Dropbox/Experiments/Tracking Method/Expand Mouse Tracking Data/')

if(!require("tidyverse")) {install.packages("tidyverse"); require("tidyverse")}
if(!require("stringr")) {install.packages("stringr"); require("stringr")}
if(!require("tibble")) {install.packages("tibble"); require("tibble")}

my_data<- read.csv(file = "New Cheating Paradigm?_July 30, 2018_13.53.csv")
## NOTE FOR STEPHANIE: Don't start/end list with delimiter 

totalPages = max(str_count(my_data$perPagePT, "PAGE BREAK"))+1
originalColumnNames <- colnames(my_data)
## separate arguments = 1. column to split 2. labels of column (finds total columns based on number of pages in survey) 3. Delimiter
df <- my_data %>% separate(perPagePT, sprintf("Page_%d", 1:(totalPages)),  sep = "PAGE BREAK", extra = "drop", fill = "right")

## Drop all other columns, only one column (with array of times) per page remains
timingArrays <- df[ , !(names(df) %in% originalColumnNames)] 

arrayColumnNames <- colnames(timingArrays)
## Loop through the column for each page and extract additional columns of data (clickAways, timeOffPage, timeOnPage )
for(i in 1:totalPages) {
  
  columnName <- names(timingArrays)[i]
  ## Separate so one there is one time per cell
  boundaryCrossings <- timingArrays %>% separate_(columnName, sprintf("Crossing_%d", 0:max(str_count(timingArrays[,i], "\\."),na.rm = T)),sep = ",", fill = "right")
  ## Drop all other columns
  boundaryCrossings <- boundaryCrossings[ , !(names(boundaryCrossings) %in% arrayColumnNames)] 
  ## Drop Crossing_0 column (since each vector starts with a ",")
  boundaryCrossings <- boundaryCrossings[,-which(names(boundaryCrossings)=="Crossing_0")] 
  ## Convert to numeric
  boundaryCrossings <- as.data.frame(sapply(boundaryCrossings, as.numeric))
  

  ## Calculate total time on the page
  timeOnPage <- apply(boundaryCrossings, 1, function(x) if(all(is.na(x))) {""} else{sum(x[x>0],na.rm = T)})
  df <- add_column(df, timeOnPage,.after = paste("Page_", i, sep = "")) 
  ## Rename newly added column
  names(df)[which(colnames(df)== paste("Page_", i, sep = ""))+1] <-  paste(paste("Page_",i,sep = ""),"_TimeOnPage", sep = "")
  
  ## Calculate total time off the page
  timeOffPage <- apply(boundaryCrossings, 1, function(x) if(all(is.na(x))) {""} else{abs(sum(x[x<0],na.rm = T))})
  df <- add_column(df, timeOffPage,.after = paste("Page_", i, sep = ""))
  ## Rename newly added column
  names(df)[which(colnames(df)== paste("Page_", i, sep = ""))+1] <-  paste(paste("Page_",i,sep = ""),"_TimeOffPage", sep = "")
  
  ## Calculate click aways (one pair = one click away)
  clickAways <-apply(boundaryCrossings, 1,function(x) if(all(is.na(x))) {""} else{ sum(!is.na(x))})
  df <- add_column(df, clickAways,.after = paste("Page_", i, sep = ""))
  ## Rename newly added column
  names(df)[which(colnames(df)== paste("Page_", i, sep = ""))+1] <-  paste(paste("Page_",i,sep = ""),"_ClickAways", sep = "")
}

## Replace NAs with blanks
df <- sapply(df, as.character)
df[is.na(df)] <- " "
as.data.frame(df)


write.csv(df,"Output.csv")