izhbannikov / gettingandcleaningdataproject

Coursera project of Getting and Cleaning data class

README

This README describes the main steps of the script run_analysis.R in order to obtain a tidy dataset.

Step 0

After setting the work directory, download the dataset.

download.file("https://d396qusza40orc.cloudfront.net/getdata%2Fprojectfiles%2FUCI%20HAR%20Dataset.zip", "Dataset.zip", method="curl")
unzip("Dataset.zip")

Step 1

Merge the training and the test sets in order to create one data set. In this case, of the test and trainding datasets can be done with a simple rbind() command. But before, you must create complete test and training datasets by adding activity and subject ids to the measurements.

## Reading the training files:
x.train <- read.table("UCI HAR Dataset/train/X_train.txt")
y.train <- read.table("UCI HAR Dataset/train/y_train.txt")
subject.train <- read.table("UCI HAR Dataset/train/subject_train.txt")
data.train <- data.frame(x.train, y.train, subject.train)

### Reading the test files:
x.test <- read.table("UCI HAR Dataset/test/X_test.txt")
y.test <- read.table("UCI HAR Dataset/test/y_test.txt")
subject.test <- read.table("UCI HAR Dataset/test/subject_test.txt")
data.test <- data.frame(x.test, y.test, subject.test)

## Merging the train and test data sets:
data <- rbind(data.train, data.test)

Step 2

Extract the measurements on the mean and standard deviation for each measurement. This is performed with ```grep`` function for the following substrings: "mean" and "std". Case - ignored.

## Reading the feature names:
features <- read.table("UCI HAR Dataset/features.txt")
##Extracting only the measurements on the mean and standard deviation for each measurement:
mean.id <- grep("mean()",features$V2, ignore.case = F) # Column Ids for means
std.id <- grep("std()",features$V2, ignore.case = F) # Column Ids for standard deviations
ids <- c(mean.id, std.id) # Column Ids for both mean and standard deviation
data.mean.std <- cbind(data$V1.1, data$V1.2, data[, ids]) # Only means and standard deviations for each measurement
names(data.mean.std) <- c("activity", "subjId", as.character(droplevels(features$V2[ids])))

Step 3

Use descriptive activity names to name the activities in the data set. This is done with loading activity lables, lowering them and replacing activity number with corresponding labels.

activity.names <- read.table("UCI HAR Dataset/activity_labels.txt")
data.mean.std$activity <- tolower(activity.names$V2[data.mean.std$activity])

Step 4

Appropriately label the data set with descriptive variable names. This is done by removing dashes and parentheses (using gsub function) from variable names.

clean.features <- names(data.mean.std) 
clean.features <- gsub("-", "", clean.features)
clean.features <- gsub("\\(\\)", "", clean.features)
names(data.mean.std) <- c(clean.features)

Step 5

From the data set in step 4, create a second, independent tidy data set with the average of each variable for each activity and each subject. This is done with ddply function from plyr library. Data was grouped by "activity" first and then by "subjectId".

tidy.dataset <- ddply(data.mean.std, c("activity","subjId"), numcolwise(mean))

Step 6

Finally, output the tidy data table into .txt file with ignoring rownames.

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

This is end of analysis.