ennpet / getdatacourseproject2

Coursera data Science specification

Getting and Cleaning Data - Course Project 2

ennpet
27. september 2015

Goal 1 - Merging the training and the test sets to create one data set.

To merge the data sets i have created helper functions to read the data from files that take the type (training or test) as a parameter and read the data in temporary variables.

library(dplyr)

## 
## Attaching package: 'dplyr'
## 
## The following objects are masked from 'package:stats':
## 
##     filter, lag
## 
## The following objects are masked from 'package:base':
## 
##     intersect, setdiff, setequal, union

rootPath <- "UCI HAR Dataset/"

read_subject <- function (type){
    read.csv(paste(rootPath, type, "/", "subject_", type, ".txt", sep = ""), header = FALSE)
}

read_X <- function (type){
    read.table(paste(rootPath, type, "/", "X_", type, ".txt", sep = ""), header = FALSE, 
    colClasses = rep("numeric", 561))
}

read_y <- function (type){
    read.csv(paste(rootPath, type, "/", "y_", type, ".txt", sep = ""), header = FALSE)
}

subject_train <- read_subject("train")
y_train <- read_y("train")
X_train <- read_X("train")

subject_test <- read_subject("test")
y_test <- read_y("test")
X_test <- read_X("test")

Then I merged training and test data sets for subjects, activities and features by rows.

subject <- rbind(subject_train, subject_test)
y <- rbind(y_train, y_test)
x <- rbind(X_train, X_test)

Goal 3 - Using descriptive activity names to name the activities in the data set

To assign descriptive activity names I have read the "activity_labels.txt" file to a data set and used this data set to factor activities with their corresponding labels.

activities <- read.csv(paste(rootPath, "activity_labels.txt",sep = ""),
                 header = FALSE, sep = " ")

y$V1 <- factor(y$V1, levels=activities$V1, labels = activities$V2 )

Goal 4- Appropriately label the data set with descriptive variable names.

To label variable names i have read the "features.txt" file to a data set and used make.names() function to assign names to variables. The original feature labels have duplicate entries and the make.names() function makes the duplicate names unique by adding suffixes.

features <- read.csv(paste(rootPath, "features.txt", sep = ""), 
                 header = FALSE, sep = " ")
names(x) <- make.names(features$V2, unique = TRUE, allow_ = TRUE)

Then I have added descriptive names to subject and activity data sets.

names(subject) <- "subject_id"
names(y) <- "activity"

Goal 1 - finishing merging the datasets

Now I have mearged the subject, activity and feature datasets by columns.

x <- cbind(subject, y, x)

Goal 2 - extracting only the measurements on the mean and standard deviation for each measurement.

x.mean.std <- select(x, subject_id, activity, contains(".mean.."), contains(".std.."))

Goal 5 - From the data set in step 4, creates a second, independent tidy data set with the average of each variable for each activity and each subject.

I have grouped the resulting data set by subject and activity and summerized it using the mean() function.

x.tidy <- x.mean.std %>%
    group_by(subject_id, activity) %>%
    summarise_each(funs(mean))

Then I have removed extra periods in the feature variable names (made by make.name() function) and saved the data set to a file.

names(x.tidy) <-(sub(".mean..", "-mean", names(x.tidy)))
names(x.tidy) <-(sub(".std..", "-std", names(x.tidy)))

write.table(x.tidy, "tidy.txt", row.names = FALSE)