Repo for Coursera course Getting and Cleaning Data by Johns Hopkins University

• run_analysis.R script
• ReadMe markdown document
• Codebook markdown document
• data folder that includes all the raw data

1. First download the data from https://d396qusza40orc.cloudfront.net/getdata%2Fprojectfiles%2FUCI%20HAR%20Dataset.zip unzip the folder and rename it as "data". Or directly download the data folder in this repo.
2. Make sure the data folder and the run_analysis.R file are both in the current working directory.
3. If you don't have tidyverse package installed, please first run install.packages("tidyverse") in RStudio.
4. Use the command source("run_analysis.R") in RStudio console.
5. You will find one output file called "tidyset.txt" in your working directory.
6. Use the command checkdata<-read.table("./tidyset.txt",header = T) in RStudio to read the file. The dataset has 180 means (30 subjects * 6 activities) of 81 features.

1. Merge the measurement set (X), the activity_label set (Y) and the subject set for both the training set and the test set respectively, then merge the training set and the test set into the set all.

2. Rename the variables in the set all by the measured features. Extract only the features with the function mean() and std() using grep() to create a new dataset all_sub. I choose to exclude the features with angle() because they are not measuring the mean, but the angle between different vectors. The following variables are selected.

• mean(): Mean value
• std(): Standard deviation
• meanFreq(): Weighted average of the frequency components to obtain a mean frequency

3. In the activity column, replace the numbers 1-6 with the descriptive labels such as "walking", "sitting" etc.

4. In the column/variable names, remove "()" and "-" and replace "mean" with "Mean", "std" with "Std".

5. Group the set all_sub by both subject and activity using the group_by function from the dplyr package. Then use the summarize_all function to find the average of each variable for each subject and each activity. The resulting dataset is named tidyset.