<<<<<<< HEAD
This project contains only one R script called run_analysis.R.
The script works in R with no need for additional packages as long as the HAR Dataset is in the working directory under the name of UCI HAR Dataset.
The HAR Dataset folder should be found at this link. In order for run_analysis.R to work, this folder must be unzipped in R's working directory under the name of UCI HAR Dataset.
As it can be read in the README.txt file of the mentioned data set, the information corresponds to several measurements of accelerometers and gyroscopes in a Samsung Galaxy S smartphone on the waists of 30 subjects performing different activities. For more details, please refer to the mentioned data set's documentation.
The first part of the script extracts all the information from different files in UCI HAR Dataset and its subfolders. Then it merges them in order to form a single data set called DS1. The script uses the read.table() function to extract the information and the cbind() and rbind() functions to merge the data sets.
The second part of the script subests DS1 in order to keep only the variables representing the mean and standard deviation for each measurement.
According to the data set's documentation, variables (features) containing -mean() and -std() in their names correspond to the measurements' mean and standard deviation respectively.
The script first finds the indexes of such variables by applying the grep() function and then subsets DS1 by those indexes. Now the DS1 data frame contains only the desired variables.
The project instructions required to apply descriptive activity names to the activity variable instead of integers, as included in the original data set.
Despite the original data set containing a file called activity_labels.txt with the label code, the renaming of the activity variable was done manually, since there are only 6 factors. The variable was saved as character. The assignments described in activity_labels.txt were kept in this step.
For Step 4, Camel case was used in order to improve readability, since the names of the variables are too long if they're to be descriptive. Although this goes against the all lowercase rule of tidy variable names, the user should consider the benefits of this practice, as suggested by TA David Hood in the Step 4 forum (see this comment).
In order to fix the variable names, the gsub() function was used several times. The description of the variables is documented in CodeBook.md within this repository and to more detail in the features_info.txt file in the HAR Dataset.
The last part of this script creates a second, independent tidy data set containing the average of each variable for each subject and each activity. The data set is stored as DS2.
The data set is generated by the aggregate() function.
More detailed information on the transformations executed by this script can be found in the CodeBook.md file in this repository.
Course project for the Getting and Cleaning Data course on Coursera.org, part of the Data Science Specialization
8bd9cd13b3daf812b53182d2ee9466c4a30c984a