run_analysis.R

### Coursera "Getting and Cleaning Data" Course, Week 4 Assignment
### Author: Sarah Gartenmann
### Date: February, 2021

### Description of Week 4 Assignment:

# This R Script is for the peer-graded assignment of the "Getting and Cleaning 
# Data" course from the Data Science Specialization offered by the John Hopkins 
# University through Coursera. As stated by the instructors of the course, "The 
# purpose of the assignment is to demonstrate your ability to collect, work with, 
# and clean a data set. The goal is to prepare tidy data that can be used for 
# later analysis". 

################################################################################
# load required packages 
library(tidyverse)
library(stringr)
library(plyr)

################################################################################
# setting the working directory and creating a new directory for the file
setwd("C:/Users/gartens1/Downloads")
if(!file.exists("./coursera_getcleandata_assignment")) {
  dir.create("./coursera_getcleandata_assignment")
}

# downloading the file (url)
fileurl <- "https://d396qusza40orc.cloudfront.net/getdata%2Fprojectfiles%2FUCI%20HAR%20Dataset.zip"
download.file(fileurl, destfile = "./coursera_getcleandata_assignment/dataset.zip")

# unzipping the zip file
unzip(zipfile = "./coursera_getcleandata_assignment/dataset.zip", 
      exdir = "./coursera_getcleandata_assignment")

################################################################################
# Assignment Part 1: merge the training and test set to create one data set

# read in features.txt file using read.table() and change the variable type to 
# character
features <- read.table("./coursera_getcleandata_assignment/UCI HAR Dataset/features.txt")
features_names <- as.character(features$V2)

# read in the 3 test datasets (subject_test.txt, X_test.txt, and y_test.txt)
# using read.table() and define the column names using rename() and 
# rename_all()
subject_test <- read.table("./coursera_getcleandata_assignment/UCI HAR Dataset/test/subject_test.txt") %>%
  rename(c("V1" = "subject_ID"))

x_test <- read.table("./coursera_getcleandata_assignment/UCI HAR Dataset/test/X_test.txt") %>%
  rename_all(~features_names)

y_test <- read.table("./coursera_getcleandata_assignment/UCI HAR Dataset/test/y_test.txt") %>%
  rename(c("V1" = "activity_ID"))

# read in the 3 training datasets (X_train.txt, y_train.txt, and subject_train)
# using read.table() and define the column names using rename() and rename_all()
subject_train <- read.table("./coursera_getcleandata_assignment/UCI HAR Dataset/train/subject_train.txt") %>%
  rename(c("V1" = "subject_ID"))

x_train <- read.table("./coursera_getcleandata_assignment/UCI HAR Dataset/train/X_train.txt") %>%
  rename_all(~features_names)

y_train <- read.table("./coursera_getcleandata_assignment/UCI HAR Dataset/train/y_train.txt") %>%
  rename(c("V1" = "activity_ID"))

# Merging
# merging the test data sets together using cbind()  
testmerged <- cbind(y_test, subject_test, x_test)

# merging the train data sets together using cbind()  
trainmerged <- cbind(y_train, subject_train, x_train)

# merging testmerged and trainmerged data sets together using rbind()
alldata <- rbind(testmerged, trainmerged)

################################################################################
# Assignment Part 2: extract only the measurements on the mean and standard 
# deviation for each measurement 

# generate the vector names_subset with the terms "mean" and "std"  
names_subset <- c("mean", "std")

# use grepl to extract all columns containing the terms from the names_subset 
# vector and save this as alldata_subset
alldata_subset <- alldata[,grepl(names_subset, names(alldata))]

# subset the columns activity_ID and activity_type from alldata and save it as 
# ID_type_subset 
ID_type_subset <- alldata[,1:2]

# merge the two data sets together and save it as the object alldata_meanstd
alldata_meanstd <- cbind(ID_type_subset, alldata_subset)

################################################################################
# Assignment Part 3: use descriptive activity names to name the activities in 
# the data set

# read in the data for the activity lables and rename the columns to activity_ID 
# and activity_name
activitylabs <- read.table("./coursera_getcleandata_assignment/UCI HAR Dataset/activity_labels.txt") %>%
  rename(c("V1" = "activity_ID",
           "V2" = "activity_name"))

# merge alldata_meanstd and activity_labs together by the column "activity_ID" 
# on all columns from the first data set (i.e. x) 
alldata_meanstd_actnames <- merge(x = alldata_meanstd, 
                                  y = activitylabs, 
                                  by = "activity_ID", 
                                  all.x = TRUE)

# reorganise the columns so that the activity_name column is placed in 2nd 
# (and not last) position 
alldata_meanstd_actnames <- alldata_meanstd_actnames[, c(1, 49, 2:48)]

################################################################################
# Assignment Part 4: appropriately label the data set with descriptive variable 
# names

# subset the column names from the original all_data 
columnnames <- colnames(alldata_meanstd_actnames)

# use str_replace_all to replace all abbreviations and include underscores 
# in between all terms 
columnnames2 <- columnnames %>%
  str_replace_all(c("Acc" = "Accelerometer_", 
                    "Gyro" = "Gyroscope_", 
                    "Mag" =  "Magnitude_",
                    "Freq" = "Frequency", 
                    "mean()" = "Mean",
                    "std" = "Standard_Deviation",
                    "BodyBody" = "Body_",
                    "Body" = "Body_",
                    "Jerk" = "Jerk_",
                    "Gravity" = "Gravity_",
                    "X" = "_X",
                    "Y" = "_Y",
                    "Z" = "_Z",
                    "MeanFrequency" = "Mean_Frequency", 
                    "Body__" = "Body_",
                    "activity_ID" = "Activity_ID",
                    "activity_name" = "Activity_Name",
                    "subject_ID" = "Subject_ID"))

# change all "t"s and "f"s to "Time_Domain" and "Frequency_Domain", respectively,
# and get rid of "()"
columnnames2 <- columnnames2 %>%
  str_replace_all(c("^t" = "Time_Domain_", 
                    "^f" = "Frequency_Domain_",
                    "[\\(\\)-]" = ""))
# Take a look to see if str_replace_all() worked like it should have 
View(columnnames2)

# Use these new labels as the column names in alldata_meanstd_actnames dataset
colnames(alldata_meanstd_actnames) <- columnnames2

################################################################################
# Assignment Part 5: from the data set in part 4, create a second, independent 
# tidy dataset with the average of each variable for each activity & each subject 

# use the group_by() function to group the dataset by subject and activity
alldata_tidy <- alldata_meanstd_actnames %>%
  group_by(Subject_ID, Activity_Name) %>%
  # calculate the mean of each column using summarise_all()
  summarise_all(.funs=mean) %>%
  ungroup() %>%
  # order the data by subject and activity_ID  
  arrange(Subject_ID, Activity_ID)

# The dataframe "alldata_tidy" looks neat and shows us the average of each variable
# for each activiry & each subject, however it does not yet fulfill all three of
# the criteria for tidy data. 
# Tidy Data means:   
  # Each variable forms a column.
  # Each observation forms a row.
  # Each type of observational unit forms a table

# create subID vector for all subject_IDs. We have 30 subject_ID (1:30), and as 
# each subject_ID performs 6 activities for each of which there are 46 measurements,
# we will have 276 (6*46) measurements per subject_ID. 
# This vector has the dimensions [1:8280]
subID <- rep(c(1:30), each = 276)

# create actID vector for all activity_IDs. We have 6 activity_ID (see "activitylabs")
# which each occur 46 times for each subject_ID. This vector has the dimensions 
# [1:8280]
actID <- rep(rep(c("Walking", 
                   "Walking Upstairs", 
                   "Walking Downstairs", 
                   "Sitting", 
                   "Standing", 
                   "Laying"), each = 46), times = 30)

# generate a new column called "measurement" which contains the column names 
# from the alldata_meanstd_actnames (columns 4:49). This character vector has 
# 46 entries. Repeat this 180 times as we need to have each measurement name 46 
# times for each activity_ID for each subject_ID. 
# This vector has the dimensions [1:8280]
measurements <- colnames(alldata_meanstd_actnames [4:49])
measurements <- rep(measurements, times = 180)

measurements <- as.data.frame(measurements)

# merge the vectors subID, actID, and measurement to create the dataframe tidy1 
# This dataframe has 8280 observations for 3 variables. Rename the columns 
tidy1 <- cbind(subID, actID, measurements) %>%
  rename(c("subID" = "Subject_ID", 
           "actID" = "Activity_ID",
           "measurements" = "Measurement_Type"))

# transpose the alldata_tidy dataframe, remove the first three rows, and unlist 
# it so that all data is in a single column called "Average_Measurement_Values"
alldata_tidy2 <- as.data.frame(t(alldata_tidy))
alldata_tidy2 <- alldata_tidy2[-(1:3),]
alldata_tidy2 <- data.frame(Average_Measurement_Value = unlist(alldata_tidy2))

#create the final dataframe "finaldata_tidy" by column binding the columns from
# "tidy1" together with those of "alldata_tidy2" and rename the columns appropriately

finaldata_tidy <- cbind(tidy1, alldata_tidy2$Average_Measurement_Value) 

finaldata_tidy <- finaldata_tidy %>%
  rename(c("alldata_tidy2$Average_Measurement_Value" = "Average_Measurement_Value",
           "Activity_ID" = "Activity_Name"))

# The final dataframe "finaldata_tidy" fulfills all criteria of tidy data!

# Create the output file "tidydata.txt" using wirte.table() 
write.table(finaldata_tidy, 
            "./coursera_getcleandata_assignment/tidydata.txt", 
            row.names = FALSE,
            col.names = TRUE,
            quote = FALSE)