### R script to run the pre-constructed, trained NN on the new data set

testdata <- read.csv("test.csv", sep=",")

# Converting factors to numbers in Sex and Cabin
testdata$Sex<-as.numeric(testdata$Sex)
testdata$Cabin<-as.numeric(testdata$Cabin)

# Approximating the ages of the NA values (linear interpolation)
testdata$Age<-na.spline(testdata$Age)
testdata$Fare<-na.spline(testdata$Fare)

# Subsetting the test data to be of the same dimensions as the train data set
testdata <- testdata[,c(2,4:7,9,10)]

# method of normalization used is the min-max scaling technique [0,1]
maxs <- apply(testdata, 2, max) 
mins <- apply(testdata, 2, min)
# scaling the data for each column and fitting the scaled values into the data frame
scaled <- as.data.frame(scale(testdata, center = mins, scale = maxs - mins))
# applying the scaled transform on the training and testing indices
testdata_ <- scaled[1:nrow(testdata),]

# finding the normalized test results for classification 
Survived.nn <- compute(nn,testdata_[,1:ncol(testdata)])  
# Using a strict threshold of 0.5 and greater for binary values of 0 & 1
Survived.nn.res <- ifelse(Survived.nn$net.result>=0.5, 1, 0)

# # comparing the two values and printing out a success percentage
# MSE.nn <- sum(abs(test_$Survived - nn.res))/nrow(test_)
# 
# #print(paste(MSE.lm,MSE.nn))
# print(paste("MSE for neural network: ",MSE.nn))