GACDProject/CodeBook at master · alexbaur/GACDProject · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
# Getting and Cleaning Data Wearable Device Project
Code Book
Alex Baur
11/19/15
true

## Project Description

Using the data from: http://archive.ics.uci.edu/ml/datasets/Human+Activity+Recognition+Using+Smartphones the task was to clean and
arrange information from the accelerometer measurements of Galaxy S II devices worn by 30 participants while doing a variety of
physical activities.

## Study design and data processing

From the UCI Machine Learning Repository linked above:

"he experiments have been carried out with a group of 30 volunteers within an age bracket of 19-48 years. Each person performed six
activities (WALKING, WALKING_UPSTAIRS, WALKING_DOWNSTAIRS, SITTING, STANDING, LAYING) wearing a smartphone (Samsung Galaxy S II) on the
waist. Using its embedded accelerometer and gyroscope, we captured 3-axial linear acceleration and 3-axial angular velocity at a
constant rate of 50Hz. The experiments have been video-recorded to label the data manually. The obtained dataset has been randomly
partitioned into two sets, where 70% of the volunteers was selected for generating the training data and 30% the test data.

The sensor signals (accelerometer and gyroscope) were pre-processed by applying noise filters and then sampled in fixed-width sliding
windows of 2.56 sec and 50% overlap (128 readings/window). The sensor acceleration signal, which has gravitational and body motion
components, was separated using a Butterworth low-pass filter into body acceleration and gravity. The gravitational force is assumed to
have only low frequency components, therefore a filter with 0.3 Hz cutoff frequency was used. From each window, a vector of features
was obtained by calculating variables from the time and frequency domain."

## Creating the tidy datafile

The tidy datafile was created through the following steps:

1. Download the data from the UCI machine learning repository.
2. Assign the test and train variables (subject, X, Y) to objects that read them as a table.
3. Combine the corresponding test and train variables.
4. Assign names to the values for X from the "features" document.
5. Combine all 3 data frames together.
6. Extract the mean and standard deviation measurements from "features" and extract the corresponding rows from the combined data frame.
7. Extract the activity labels from the "activity_labels" file and implement in the "Activity_type" column.
8. Alter the names for the various measurements to be less abbreviated and easier to understand.
9. Aggregate and write the table to "tidydata.txt".

## Cleaning of the data

The cleaning script takes the X_test/X_train, Y_test/Y_train, and subject_test/subject_train files and combines them to form 3 data
frames. It then renames the variables (including all of the variables for X) and merges the entire data set together. AFter this, it
decodes the activity data collected by replacing the names for each column with a simplified descriptor that can be more readily
understood. Finally, it arranges by the mean so that trends can be observed. Please see the readme in this Repo for further details.

## Variables

The tidydata text file is comprised of 88 columns of 80 rows. It identifies the subject the data was taken from (subject_number, 30
participants total), the type of activity they were doing when the recording was taken (Activity_type, 6 different types), and 86
variables concerning the specific measurements made by the gyroscope and accelerometer in the Galaxy S II device (86 based on x, y, z
coordinate reports).

### Subject_number

Describes which participant the information is coming from. They are represented by an integer from 1-30 with no unit.

### Activity_type

Describes what activity the subject was performing when the readings were taken. There are 6 levels corresponding to the following:

1 WALKING
2 WALKING_UPSTAIRS
3 WALKING_DOWNSTAIRS
4 SITTING
5 STANDING
6 LAYING

These variables were retrieved from the activity_labels document in the zip file found in the sources section below. Each of these
is a unitless factor.

### Movement Measurements

These are found in columns 3:88 of the tidy data file. They are all numeric and correspond to the means and standard deviations of the
x, y, and z coordinates for accelerometers and gyroscopes present in the Galaxy S II device. These are then distinguished by time,
and angle denotations which are also then further broken down by magnitude and jerk measurements. These variables are listed below and
further explanation of them can be found in the README in the zipfile listed in the sources section.

"TimeBodyAccelerometerMean()-X"                              "TimeBodyAccelerometerMean()-Y"
"TimeBodyAccelerometerMean()-Z"                              "TimeBodyAccelerometerStdDev()-X"
"TimeBodyAccelerometerStdDev()-Y"                            "TimeBodyAccelerometerStdDev()-Z"
"TimeGravityAccelerometerMean()-X"                           "TimeGravityAccelerometerMean()-Y"
"TimeGravityAccelerometerMean()-Z"                           "TimeGravityAccelerometerStdDev()-X"
"TimeGravityAccelerometerStdDev()-Y"                         "TimeGravityAccelerometerStdDev()-Z"
"TimeBodyAccelerometerJerkMean()-X"                          "TimeBodyAccelerometerJerkMean()-Y"
"TimeBodyAccelerometerJerkMean()-Z"                          "TimeBodyAccelerometerJerkStdDev()-X"
"TimeBodyAccelerometerJerkStdDev()-Y"                        "TimeBodyAccelerometerJerkStdDev()-Z"
"TimeBodyGyroscopeMean()-X"                                  "TimeBodyGyroscopeMean()-Y"
"TimeBodyGyroscopeMean()-Z"                                  "TimeBodyGyroscopeStdDev()-X"
"TimeBodyGyroscopeStdDev()-Y"                                "TimeBodyGyroscopeStdDev()-Z"
"TimeBodyGyroscopeJerkMean()-X"                              "TimeBodyGyroscopeJerkMean()-Y"
"TimeBodyGyroscopeJerkMean()-Z"                              "TimeBodyGyroscopeJerkStdDev()-X"
"TimeBodyGyroscopeJerkStdDev()-Y"                            "TimeBodyGyroscopeJerkStdDev()-Z"
"TimeBodyAccelerometerMagnitudeMean()"                       "TimeBodyAccelerometerMagnitudeStdDev()"
"TimeGravityAccelerometerMagnitudeMean()"                    "TimeGravityAccelerometerMagnitudeStdDev()"
"TimeBodyAccelerometerJerkMagnitudeMean()"                   "TimeBodyAccelerometerJerkMagnitudeStdDev()"
"TimeBodyGyroscopeMagnitudeMean()"                           "TimeBodyGyroscopeMagnitudeStdDev()"
"TimeBodyGyroscopeJerkMagnitudeMean()"                       "TimeBodyGyroscopeJerkMagnitudeStdDev()"
"FrequencyBodyAccelerometerMean()-X"                         "FrequencyBodyAccelerometerMean()-Y"
"FrequencyBodyAccelerometerMean()-Z"                         "FrequencyBodyAccelerometerStdDev()-X"
"FrequencyBodyAccelerometerStdDev()-Y"                       "FrequencyBodyAccelerometerStdDev()-Z"
"FrequencyBodyAccelerometerMeanFreq()-X"                     "FrequencyBodyAccelerometerMeanFreq()-Y"
"FrequencyBodyAccelerometerMeanFreq()-Z"                     "FrequencyBodyAccelerometerJerkMean()-X"
"FrequencyBodyAccelerometerJerkMean()-Y"                     "FrequencyBodyAccelerometerJerkMean()-Z"
"FrequencyBodyAccelerometerJerkStdDev()-X"                   "FrequencyBodyAccelerometerJerkStdDev()-Y"
"FrequencyBodyAccelerometerJerkStdDev()-Z"                   "FrequencyBodyAccelerometerJerkMeanFreq()-X"
"FrequencyBodyAccelerometerJerkMeanFreq()-Y"                 "FrequencyBodyAccelerometerJerkMeanFreq()-Z"
"FrequencyBodyGyroscopeMean()-X"                             "FrequencyBodyGyroscopeMean()-Y"
"FrequencyBodyGyroscopeMean()-Z"                             "FrequencyBodyGyroscopeStdDev()-X"
"FrequencyBodyGyroscopeStdDev()-Y"                           "FrequencyBodyGyroscopeStdDev()-Z"
"FrequencyBodyGyroscopeMeanFreq()-X"                         "FrequencyBodyGyroscopeMeanFreq()-Y"
"FrequencyBodyGyroscopeMeanFreq()-Z"                         "FrequencyBodyAccelerometerMagnitudeMean()"
"FrequencyBodyAccelerometerMagnitudeStdDev()"                "FrequencyBodyAccelerometerMagnitudeMeanFreq()"
"FrequencyAccelerometerAccelerometerJerkMagnitudeMean()"     "FrequencyAccelerometerAccelerometerJerkMagnitudeStdDev()"
"FrequencyAccelerometerAccelerometerJerkMagnitudeMeanFreq()" "FrequencyAccelerometerGyroscopeMagnitudeMean()"
"FrequencyAccelerometerGyroscopeMagnitudeStdDev()"           "FrequencyAccelerometerGyroscopeMagnitudeMeanFreq()"
"FrequencyAccelerometerGyroscopeJerkMagnitudeMean()"         "FrequencyAccelerometerGyroscopeJerkMagnitudeStdDev()"
"FrequencyAccelerometerGyroscopeJerkMagnitudeMeanFreq()"     "Angle(T-BodyAccelerometerMean,Gravity)"
"Angle(T-BodyAccelerometerJerkMean),GravityMean)"            "Angle(T-BodyGyroscopeMean,GravityMean)"
"Angle(T-BodyGyroscopeJerkMean,GravityMean)"                 "Angle(X,GravityMean)"
"Angle(Y,GravityMean)"                                       "Angle(Z,GravityMean)"

## Sources

https://d396qusza40orc.cloudfront.net/getdata%2Fprojectfiles%2FUCI%20HAR%20Dataset.zip
http://archive.ics.uci.edu/ml/datasets/Human+Activity+Recognition+Using+Smartphones