ass8

# Step 0: Start; Getting the starting time
cat("\nSTART\n")
startTime = proc.time()[3]
startTime



# Step 1: Include Library
cat("\nStep 1: Library Inclusion")



# Step 2: Variable Declaration and Seed Setting
cat("\nStep 2: Variable Declaration and Seed Setting")
modelName <- "linear"
modelName
InputDataFileName="regressionDataSet.csv"
InputDataFileName

seed=sample(99999:999999, 1)
seed
set.seed(seed)



# Step 3: Data Division for train/test
cat("\nStep 3: Data Division")
training=50      # define percentage



# Step 4: Load the data from file
cat("\nStep 4: Loading the data")
dataset <- read.csv(InputDataFileName)    # Read the datafile
head(dataset)   # Show Top 6 records
nrow(dataset)   # Show number of records
names(dataset)  # Show fields names or columns names



# Step 5: Count total number of observations/rows.
cat("\nStep 5: Counting dataset")
totalDataset <- nrow(dataset)
totalDataset



# Step 6: Choose Target variable
cat("\nStep 6: Choose Target Variable")
target  <- names(dataset)[1]   # i.e. RMSD
target



# Step 7: Choose inputs Variables
cat("\nStep 7: Choose Inputs Variable")
inputs <- setdiff(names(dataset),target)
inputs
length(inputs)


selectedInputs <- inputs


# Step 8: Select Training Data Set ( Select random indices )
cat("\nStep 8: Select training dataset")
trainSample <- sample(totalDataset, totalDataset * training/100)
head(trainSample)     # Show Top 6 indecies
length(trainSample)   # Show number of indecies

trainDataset <- dataset[trainSample,c(selectedInputs, target)]
head(trainDataset)    # Show Top 6 records
nrow(trainDataset)    # Show number of train Dataset



# Step 9: Select Testing Data Set
cat("\nStep 9: Select testing dataset")
testSample <- setdiff(seq_len(nrow(dataset)), trainSample)
head(testSample)      # Show Top 6 indecies
length(testSample)    # Show number of indecies

testDataset <- dataset[testSample,c(selectedInputs, target)]
head(testDataset)
nrow(testDataset)




# Step 10: Regression Model Building: Training
cat("\nStep 10: Model Building -> ", modelName)
formula <- as.formula(paste(target,"~",paste(c(selectedInputs),collapse = "+")))
formula

model   <- lm(formula, trainDataset)
model
summary(model)


# Step 11: Prediction (Testing)
cat("\nStep 11: Prediction using -> ", modelName)
Predicted <- predict(model, testDataset)
head(Predicted)



# Step 12: Extracting Actual
cat("\nStep 12: Extracting Actual")
Actual <- as.double(unlist(testDataset[target]))
head(Actual)




# Step 13: Model Evaluation
cat("\nStep 13: Model Evaluation")

# Step 13.1: Correlation
r <- cor(Actual,Predicted )
r <- round(r,2)
r

# Step 13.2: Accuracy
accuracy <- mean(abs(Actual-Predicted) <=1)
accuracy <- round(accuracy,4) *100
accuracy

# Step 13.5: Total Time
totalTime = proc.time()[3] - startTime
totalTime

# Step 13.6: Scatter Plot
png(filename=paste(modelName,"-ScatterPlot.png",sep=''))

plot(Actual,Predicted,main=paste("Actual Vs Predicted\n",modelName),xlab="Predicted", ylab="Actual")#, pch=19)

abline(lm(Actual ~ Predicted,),col="White")

dev.off()