---title: "README"author: "Nikolos Peyralans"date: "March 4, 2016"output: html_document---##STEP I: Merging the data setsLoad required packages:```{r, echo=FALSE}library(dplyr)library(plyr)library(data.table)```#####Activity data```{r}activity_test <- read.table("test/y_test.txt", header = FALSE)activity_train <- read.table("train/y_train.txt", header = FALSE)```#####Subject data```{r}subject_test <- read.table("test/subject_test.txt", header = FALSE)subject_train <- read.table("train/subject_train.txt", header = FALSE)```#####Features file```{r}features_test <- read.table("test/x_test.txt", header = FALSE)features_train <- read.table("train/x_train.txt", header = FALSE)```#####Merge data sets based on rows```{r}subject_data <- rbind(subject_train, subject_test)activity_data <- rbind(activity_train, activity_test)features_data <- rbind(features_train, features_test)```#####Name the columns```{r}names(subject_data) <- "subject"names(activity_data) <- "activity"features_data_names <- read.table("features.txt",head=FALSE)names(features_data) <- features_data_names$V2```#####Merge all of this data with columns "subject", "activity", and "features"```{r}subject_activity_data <- cbind(subject_data, activity_data)all_data <- cbind(features_data, subject_activity_data)```##STEP II: Extracting mean and standard deviation records#####Use regular expressions to extract mean and standard deviation records```{r}mean_sd_features_names <- features_data_names$V2[grep("mean\\(\\)|std\\(\\)", features_data_names$V2)]```#####Prepare data frame for values```{r}selected_names <- c(as.character(mean_sd_features_names), "subject", "activity")```#####Make the data frame with only mean and sd values```{r}mean_sd_data <- subset(all_data, select=selected_names)```##STEP III: Naming#####Change activity labels```{r}mean_sd_data$activity[mean_sd_data$activity == 1] <- "walking"mean_sd_data$activity[mean_sd_data$activity == 2] <- "walking_upstairs"mean_sd_data$activity[mean_sd_data$activity == 3] <- "walking_downstairs"mean_sd_data$activity[mean_sd_data$activity == 4] <- "sitting"mean_sd_data$activity[mean_sd_data$activity == 5] <- "standing"mean_sd_data$activity[mean_sd_data$activity == 6] <- "laying"```#####Factorize```{r}mean_sd_data$activity <- as.factor(mean_sd_data$activity)```#####Change the names of the data```{r}names(mean_sd_data) <- gsub("^t", "time", names(mean_sd_data))names(mean_sd_data) <- gsub("^f", "frequency", names(mean_sd_data))names(mean_sd_data) <- gsub("Acc", "Accelerometer", names(mean_sd_data))names(mean_sd_data) <- gsub("Gyro", "Gyroscope", names(mean_sd_data))names(mean_sd_data) <- gsub("Mag", "Magnitude", names(mean_sd_data))names(mean_sd_data) <- gsub("BodyBody", "Body", names(mean_sd_data))```##STEP IV: Making an independent data set with average of each variable for each activity and each subject```{r}mean_sub_act <- aggregate(. ~subject + activity, mean_sd_data, mean)```#####Order it by subject```{r}mean_sub_act <- mean_sub_act[order(mean_sub_act$subject, mean_sub_act$activity),]```#####Write the new data set to a file```{r}write.table(mean_sub_act, file = "mean_subject_activity.txt", row.name=FALSE)```