ivogeorg / harus-project

Coursera getdata-013 course project on data for Human Activity Recognition Using Smartphones.

harus-project

Coursera getdata-013 course project on data for Human Activity Recognition Using Smartphones. The project involves merging and transforming an original dataset for machine learning for recodnition by mobile devices of standard human activities by measuring signals from the device's built-in accelerometer and gyroscopes.

Original dataset

The original dataset can be found on the UCI Machine Learning Repository site. The documentation of the original dataset, along with in-depth explanation of the experimental settings, can be found in the data directory. A good place to start is the supplied README.txt.

Transformations

The transformations comprising the project are performed by the run_analysis.R script. The script has inline comments marking out all each major step. In summary, the original dataset, which is split in a 70:30 ratio into training and test sets, is:

1. merged, both horizontally, across the train-test dimension, and vertically, along the subject-features-label dimension;

2. filtered, to retain only the columns for the mean and standard deviation values for 33 of the original feature variables, which are described in the file features_info.txt;

3. summarized, to show each of the variables averaged over each activity for each subject; and

4. cleaned up, so that variable names and factors are human readable

Resulting dataset

The resulting dataset is saved in plain-text space-delimited file subj_acty_avg.txt in the data directory. It can be imported into R/RStudio with read.table() with the option header=TRUE.

The dataset is tidy per the requirements of Week 1 of the course getdata-013. It is in the wide-data format.

The data dictionary codebook.md contains descriptions of the variables in the new dataset.