cscAssign1.R

working_dir<-"\\afs\\unity.ncsu.edu\\users\\e\\efarhan\\csc-591"

file1<-"student_prb_clean.csv"
file2<-"student_step_clean.csv"


### concatenate path location to filename
input_stdProb<-paste(working_dir,"\\",file1, sep="")
input_stdStep<-paste(working_dir,"\\", file2,sep="")

###read .csv files
std_prb<-read.table(file1, sep=",",header=T,check.names = FALSE)
std_step<-read.table(file2, sep=",",header=T, check.names = FALSE)


## Extract columns "Anon Student Id", "Correct First Attempts" from std_prb file
col_std_prb<-c("Anon Student Id", "Correct First Attempts")#"Anon Student Id", "Correct First Attempts"
subset_std_prob<-std_prb[,col_std_prb]


## Extract columns "Anon Student Id", "Correct First Attempts" from std_step file
col_std_step<-c("Anon Student Id", "Step Duration (sec)")#"Anon Student Id", "Step Duration"
subset_std_step<-std_step[, col_std_step]


##get all distinct student names from subset_std_step (subset_std_prob)
std_names<-as.character(unique(subset_std_step[,1]))

### create an empty data frame to fill up std_name, total corr attempts, and total step duration
b22<-data.frame(name_of_std=character(),
                total_corr_attempts=integer(), 
                total_step_duration=double(), 
                stringsAsFactors=FALSE)


for(i in 1:length(std_names))
{
  ### get a std name, and get row indices of correct first attempts and step duration for that student from two different data frames 
  curr_std<-std_names[i]
  get_attemt_rows<-which(subset_std_prob[,1]==curr_std)
  get_step_rows<-which(subset_std_step[,1]==curr_std)
  
  ###### Extract the values of correct first attempts from row indices
  get_attemt_row_extract<-subset_std_prob[get_attemt_rows,]
  get_attemt_val<-get_attemt_row_extract[,2]
  
  ###### Extract the values of step duration(sec) from row indices
  get_step_row_extract<-subset_std_step[get_step_rows,]
  get_step_dur<- get_step_row_extract[,2]
  
  ###sume up values of total correct first attempts and step duration(sec), and add to the new data frame created, b22
  b22[i,1]<-curr_std
  b22[i,2]<-sum(get_attemt_val)
  b22[i,3]<-sum(get_step_dur)
}


##### Find outliers of extracted two measures and compare with and without outlier values
to_remove_again<-c()

### for each column find outliers identify corresponding rows, and add them to_remove vector 
## Exclude first column, this contains student names
col_b22<-names(b22)[-1]

for (Var in col_b22) {
  
  outliers<-boxplot.stats(b22[,Var])$out
  
 # cat("out ", outliers, "\n")
  outlier_rows<-c()
  j<-1
  while( j<=length(outliers)){
     outlier_rows<-c(outlier_rows, which(b22[,Var]==outliers[j]))
     j<-j+1
  }
  
  to_remove_again<-c(to_remove_again, outlier_rows)
}

### View the two basic per student performance statistics, with outliers
View(b22)


###remove outlier rows, if outlier_rows>0
b22_clean<-b22

if(length(to_remove_again)>0)
   b22_clean<-b22[-to_remove_again,]



####This code segment outputs statistics with outliers
cat("First Corr. Attempts Step Duration (sec)")
cat(" mean with outliers", mean(b22[,2]), " ", mean(b22[,3]), "\n")
cat(" median with outliers", median(b22[,2]), " ", median(b22[,3]), "\n")
cat(" max with outliers", max(b22[,2]), " ", max(b22[,3]),"\n")
cat("  min with outliers", min(b22[,2]), " ", min(b22[,3]), "\n")


####This code segment outputs statistics without outliers
if(length(to_remove_again)>0){
  cat(" mean w/o outliers", mean(b22_clean[,2]), " ", mean(b22_clean[,3]), "\n")
  cat(" median w/o outliers",median(b22_clean[,2]), " ", median(b22_clean[,3]), "\n")
  cat(" median w/o outliers", max(b22_clean[,2]), " ", max(b22_clean[,3]), min(b22_clean[,2]), " ", min(b22_clean[,3]),  "\n")
}


###Plotting for Correct First Attempts
attemps<-b22_clean[,2]
z.norm<-(attemps-mean(attemps))/sd(attemps)## standardize points
qqnorm(z.norm) ## drawing the QQplot 
abline(0,1, col="blue") ## drawing a 45-degree reference line
hist(attemps,main="Histogram of Correct First Attempts")


###Plotting for Correct First Attempts
durations<-b22_clean[,3]
z.norm<-(durations-mean(durations))/sd(durations)
qqnorm(z.norm) ## drawing the QQplot 
abline(0,1, col="red") ## drawing a 45-degree reference line
hist(durations,main="Histogram of Total Step Duration (sec)")