Dataset downloaded from
into a directory named UCI HAR Dataset
The script run_analysis.R assumes that it is being run in the directory holding the dataset
- Input: UCI HAR Dataset. See the data description below.
- Output: cleanData.txt This file contains data from the original training and test data files. Only the mean and standard deviation features are retained. Descriptive labels are provided for subject activities. summaryData.txt This file contains the mean of the features in cleanData.txt aggregated by subject and activity. Output files can be read with read.table("file_name")
For every one of 30 subjects, a 561-feature vector was created capturing data for each of 6 activities.
For the course summary data set, I have extracted the data labeled "mean()" and "std()" from the test and training sets for each of the 33 features listed below. (Data labeled XYZ consist of 3 data points)
- tBodyAccXYZ
- tGravityAccXYZ
- tBodyAccJerkXYZ
- tBodyGyroXYZ
- tBodyGyroJerkXYZ
- tBodyAccMag
- tGravityAccMag
- tBodyAccJerkMag
- tBodyGyroMag
- tBodyGyroJerkMag
- fBodyAccXYZ
- fBodyAccJerkXYZ
- fBodyGyroXYZ
- fBodyAccMag
- fBodyAccJerkMag
- fBodyGyroMag
- fBodyGyroJerkMag
Resulting variables, then are labeled as (e.g.) tBodyAccXMean or fBodyGyroMagStd
Although the following are listed as "Means", they had no matching standard deviation so they were not included in the analysis set.
- gravityMean
- tBodyAccMean
- tBodyAccJerkMean
- tBodyGyroMean
- tBodyGyroJerkMean
More details on each of the features in the data set can be found in UCI HAR Dataset/features_info.txt