daviddpark / gcdataassignment

Assignment files for Coursera course "Getting and Cleaning Data"

GCDataAssignment

Assignment files for Coursera course "Getting and Cleaning Data"

Working with Human Activity Recognition Using Smartphones Data Set

The dataset is geared toward a machine learning exercise, splitting the collected data into two separate datasets for training and testing. As the dataset is further broken out into separate files isolating the observed subject and the observed activity from the data collected, it is necessary to consolidate the individual files into a dataset, then combine the testing and training data into a single dataset for aggregation.

The course instructor has assigned us to do the above merge and then perform the aggregation only on the fields pertaining to mean and standard deviations of the observed data. The subset of fields from the original dataset are identified in the code book.

Assignment Details and the steps of the run_analysis.R script

Create one R script called run_analysis.R that does the following. 0. Retrieve the dataset if necessary. First the script checks for the availability of the dataset. If it is not present, the script will download and extract the dataset.

1. Merges the training and the test sets to create one data set.

   As the dataset does not contain variable names, but provides a features.txt file that contains the variable names in the order the columns appear in the dataset, the script reads in the file, and as it reads in the x_test and x_train data files, it uses the features as column names. Activity files are also imported and combined from Y_test.txt and Y_train.txt Subject files are also imported and combined from subject_test.txt and subject_train.txt. Note that when combining all_data, activities, and subjects, it is vital that the same order is used (test, then train) in order to correlate the subject and activity with the observed data.

2. Extracts only the measurements on the mean and standard deviation for each measurement. The script assembles a list of column names where the label for a column contains the text "mean" or "std", this is then used in step 5 to create a tidy data set.

3. Uses descriptive activity names to name the activities in the data set The original data set provides a file containing the activity descriptions that correspond to the ids representing the activity in the Y_train.txt and Y_test.txt files. The labeled_activities data frame is created using this mapping from the id to the String representation.

4. Appropriately labels the data set with descriptive variable names. The script modifies the original labels, taken from the features.txt file, to create column names that are a bit more meaningful. Based on the original documentation of the data set, the script prefixes variables that start with "t" to "Time", "f" to "Frequency", "Acc" is expanded to "Accel", and in order to preserve the "CamelCasing" of the variable names, the script also capitalizes the "mean" and "std" substrings.

5. Creates a second, independent tidy data set with the average of each variable for each activity and each subject. The script leverages the reshape2 library to obtain a "molten" data set with only the appropriate columns, using the cols list as described in #2 above. When the molten dataset is recast, the mean function is used to identify the average of the values for each of the desired columns.

Since the script created a data frame containing the original labels as well as the new label, the script is able to use that data frame to dynamically create the CODEBOOK.md file.