Getting and Cleaning Data- Final Project

1 - Downdloading the raw dataset from the provided link

downloadingdataset.R

Download and unzip the dataset:

filename <- "getdata_dataset.zip"

if (!file.exists(filename)){
  fileURL <- "https://d396qusza40orc.cloudfront.net/getdata%2Fprojectfiles%2FUCI%20HAR%20Dataset.zip "
  download.file(fileURL, filename, method="curl")
}  
if (!file.exists("UCI HAR Dataset")) { 
  unzip(filename) 
}

2 - Apply the corresponding transformations to get a tidy dataset (script )

run-analysis.R

Loading activity datasets

testActivity  <- read.table("UCI HAR Dataset/test/Y_test.txt" , header = FALSE)
trainActivity <- read.table("UCI HAR Dataset/train/Y_train.txt", header = FALSE)

Loading subject datasets

testSubject  <- read.table("UCI HAR Dataset/test/subject_test.txt", header = FALSE)
trainSubject <- read.table("UCI HAR Dataset/train/subject_train.txt", header = FALSE)

Loading features datasets

testFeatures  <- read.table("UCI HAR Dataset/test/X_test.txt", header = FALSE)
trainFeatures <- read.table("UCI HAR Dataset/train/X_train.txt", header = FALSE)

Loading labels

labels <- read.table("UCI HAR Dataset/activity_labels.txt", header = FALSE)

Loading features names

featuresNames <- read.table("UCI HAR Dataset/features.txt", head=FALSE)

Combining the testing and training datasets

activity <- rbind(trainActivity, testActivity)
subject <- rbind(trainSubject, testSubject)
features <- rbind(trainFeatures, testFeatures)

Aliasing activity to assing the label a human readable category

activity$V1 <- factor(activity$V1, levels = as.integer(labels$V1), labels = labels$V2)

Renaming columns for activity and subject

names(activity)<- c("activity")
names(subject)<-c("subject")

Renaming features

names(features)<- featuresNames$V2

Selecting just those columns with mean and standard deviation information

meanstdev<-c(as.character(featuresNames$V2[grep("mean\\(\\)|std\\(\\)", featuresNames$V2)]))

Subseting the dataset with the mean and std columns

featuresSubset<-subset(features,select=meanstdev)

Combining data sets: subject, activity and features

subjectactivity <- cbind(subject, activity)
finaldataset <- cbind(featuresSubset, subjectactivity)

Renaming time and frequency

names(finaldataset)<-gsub("^t", "time", names(finaldataset))
names(finaldataset)<-gsub("^f", "frequency", names(finaldataset))

Creating a new data set with subject and activity means

suppressWarnings(cleandata <- aggregate(finaldataset, by = list(finaldataset$subject, finaldataset$activity), FUN = mean))
colnames(cleandata)[1] <- "Subject"
names(cleandata)[2] <- "Activity"

Removing avg and stdev for non-aggregated sub and act columns

cleandata <- cleandata[1:68]

Writing our tidy data set into a text file

write.table(cleandata, file = "tidydataset.txt", row.name = FALSE)