###Introduction First of all, I would briefly explain the files made for course project assignment for "Getting and Cleaning Data"
Then, the explanation for the scrpit follows.

###This Github repo includes run_analysis.R, README.md, and CodeBook.md. According to the instruction,
run_analysis.R was made to output the text file from original data saved in the working directory.
README.md is this file.
CodeBook.md. contains information about the variables which appears in the script.

###Script. Actually, original script contains comments (starting with #) explaining how script works. Thus, please refer to the following script with comment.

#load data from files
test<- read.table("X_test.txt",header=F) testlabel<- read.table("y_test.txt",header=F) train<- read.table("X_train.txt",header=F) trainlabel<- read.table("y_train.txt",header=F) feature<- read.table("features.txt",header=F) actlabel<- read.table("activity_labels.txt",header=F) subjecttest<- read.table("subject_test.txt",header=F) subjecttrain<- read.table("subject_train.txt",header=F)

#task1##################################################
#Merging the training and the test sets
combine<-rbind(test,train)

#task2##################################################
#labeling the data set with descriptive variable names.
colnames(combine)<-feature[,2]

#making a vector called "labels" for extracting only the measurements on the mean and standard deviation for each measurement.

#extarcting feature including "mean"
mean<-grep("\-mean", feature[,2])

#extarcting feature including "meanfreq"
meanfreq<-grep("\-meanFreq", feature[,2])

#extarcting feature including "mean" but excluding "meanfreq"
cleanmean<-setdiff(mean, meanfreq)

#extarcting feature including "std"
std<-grep("\-std", feature[,2])

#combining those features to make a vector called "labels" and conducting sorting as well.
labels<-sort(c(cleanmean,std))

#slicing by using vector created above
combineavesd<-combine[,labels]

#task3##################################################
#Merging the labels for training and the test sets
combinelabel<-rbind(testlabel,trainlabel)

#using information from activity label file, converting numeric activity labels into the #discriptive activity names
for (i in 1:length(actlabel[,1])) {
combinelabel[,1]<-sub( actlabel[i,1], actlabel[i,2], combinelabel[,1])
}

activity_labeled_combineavesd<-cbind(combinelabel,combineavesd)

#task4##################################################
#In order to label the data set with descriptive variable names,
#original variable names were trasnformed into more descriptive ones

name<-colnames(activity_labeled_combineavesd)
name<-gsub( "V1", "Activity_name", name)
name<-gsub( "Acc", "Acceleration", name)
name<-gsub( "std", "SD", name)
name<-gsub( "Gyro", "Angular_velocity", name)
name<-gsub( "Mag", "Magnitude", name)
name<-gsub( "std", "SD", name)
name<-gsub( "X", "X_axis_signal", name)
name<-gsub( "Y", "Y_axis_signal", name)
name<-gsub( "Z", "Z_axis_signal", name)
name[2:41]<-sub( "t", "Time_domain", name[2:41])
name[42:67]<-sub( "f", "Frequency_domain", name[42:67])
colnames(activity_labeled_combineavesd)<-name

#task5##################################################
#Merging the training and the test subject information
subjects<-rbind(subjecttest,subjecttrain)

#Merging the dataset from task4 and subject information

data5<-cbind(subjects,activity_labeled_combineavesd)
colnames(data5)<-sub( "V1", "Subject", colnames(data5))

m <- data.frame(matrix(nrow=length(unique(data5[,1]))*length(unique(data5[,2])),ncol = 68))
colnames(m)<-colnames(data5)

l=1
for (i in unique(data5[,1])) {
for (j in unique(data5[,2])) {
m[l,1]=i
m[l,2]=j
l=l+1 }
}

p=1
for (i in unique(data5[,1])) {
for (j in unique(data5[,2])) {
m[p,3:68]=colMeans(data5[(data5[,1]==i & data5[,2]==j),3:68])
p=p+1
}
}

write.table(m,"week3.txt",row.name=FALSE,sep=" ")