This repository contains the final project of the course Getting and Cleaning Data [final project] (https://class.coursera.org/getdata-003/human_grading/view/courses/972136/assessments/3/submissions).
The analysis is performed by a single R script that loads the train and test data of the experiment, and then merges them into a single dataset. The script simplifies the merged dataset by keeping only the mean and standard value of each phisical measurement, and labels them according to their definitions, given in the file features.txt and features_info.txt (see the [dataset repository] (http://archive.ics.uci.edu/ml/datasets/Human+Activity+Recognition+Using+Smartphones) for more information about the experiment setup) and read CodeBook.md for more information about the analysis.
Finally, the script averages the extracted values by activity and subject, keeping a single row for each activity-subject pair. The resulting dataset is saved as in a new file (named clean_dataset.txt) in the current folder.
Note that in order for this analysis to run properly the original [dataset] (http://archive.ics.uci.edu/ml/machine-learning-databases/00240/UCI%20HAR%20Dataset.zip) should be located in the same folder, that is "./UCI HAR Dataset/"