glasswm/GetData_CourseProject

Instuctions

• Download the Human Activity Recognition Using Smartphones data set and extract. Suppose you get the xxx/UCI HAR Dataset directory.
• Download the run_analysis.R script to the same directory xxx.
• Change current directory to the xxx.
• Run Rscript run_analysis.R
• The tidy dataset should get created in the current directory as res.txt
• Code book for the tidy dataset is available here

analysis of the run_analysis.R script

1. The script first define a function to get descriptive activity names.
   eg.
      input: (1, 3, 2)
      output: ("WALKING", "WALKING_DOWNSTAIRS", "WALKING_UPSTAIRS")

getActivity <- function(x) {
  activity <- c("WALKING", "WALKING_UPSTAIRS", "WALKING_DOWNSTAIRS", "SITTING", "STANDING", "LAYING")
  res <- c()
  for (i in x) {
    res <- c(res, activity[i])
  }
  res
}

2. read raw data

x_train <- read.table("respo/GetData_CourseProject/UCI HAR Dataset/train/X_train.txt", header=FALSE, stringsAsFactors=FALSE)
y_train <- read.table("respo/GetData_CourseProject/UCI HAR Dataset/train/y_train.txt", header=FALSE, stringsAsFactors=FALSE)
subject_train <- read.table("respo/GetData_CourseProject/UCI HAR Dataset/train/subject_train.txt", header=FALSE, stringsAsFactors=FALSE)
x_test <- read.table("respo/GetData_CourseProject/UCI HAR Dataset/test/X_test.txt", header=FALSE, stringsAsFactors=FALSE)
y_test <- read.table("respo/GetData_CourseProject/UCI HAR Dataset/test/y_test.txt", header=FALSE, stringsAsFactors=FALSE)
subject_test <- read.table("respo/GetData_CourseProject/UCI HAR Dataset/test/subject_test.txt", header=FALSE, stringsAsFactors=FALSE)

3. use cbind to bind the subject person, activity label and the observation.
   use rbind to bind the train data and test data.

train <- cbind(subject_train, y_train, x_train)
test <- cbind(subject_test, y_test, x_test)
data <- rbind(train, test)

4. extract only the measurements on the mean and standard deviation

features <- read.table("respo/GetData_CourseProject/UCI HAR Dataset/features.txt", stringsAsFactors=FALSE)
mean_std <- c(grep("mean\\(\\)", features$V2), grep("std\\(\\)", features$V2))
mean_std <- mean_std + c(2)
mean_std <- c(1, 2, mean_std) 
mean_std <- sort(mean_std)
data <- data[mean_std]

5. uses descriptive activity names to name the activities in the data set

data[,2] <- getActivity(data[,2])

6. labels the data set with descriptive variable names

names <- c("subject", "activity", features$V2)[mean_std]
colnames(data) <- names

7. creates a second, independent tidy data set with the average of each variable for each activity and each subject. Write the data set to file called "res.txt"

tt <- aggregate(data[,3:68], list(data$subject, data$activity), mean)
names(tt)[1:2] = c("subject", "activity")
write.table(tt, file="respo/GetData_CourseProject/res.txt", quote=FALSE, row.names=FALSE)