Getting-and-Cleaning-Data-Project

All code for analysis is contained in 1 script, run_analysis.R

Part 1 code

The first section of code downloads the required documents from the url. An if statement is included so that files do not get redownloaded if they are already present. The Train dataset is with the 571 feature vector is combined with the participant ID and the physical activity identifier. This dataframe is organized with participant ID in the first column, the physical activity identifier in the second column, and the feature vector in the subsequent columns. The same was done for the Test dataset. These were then combined using rbind to make one large dataset with data from all the participants.

Part 2 code

The next section reads in the features.txt file which has all of the variable names for the gyroscope and accelerometer data. Using the grepl function, a logical vector of names that contain either mean or std is created. Using this, the combined data from the previous step is filtered to only contain features in accordance with the logical vector. This trimmed data is then joined with the participant ID column and physical activity column

Part 3 code

The activity_labels.txt file is read in and is used to replace the numeric representations for physical activities with strings. Underscores were removed from the strings to eliminate an downstream errors caused by including non alphanumeric characters.

Part 4 code

This part used the feature labels selected from part 2. First a vector of characters to be removed was created. Then a loop was used in conjunction with the gsub function to remove these characters. Following this, t and f were replaced with time and freq, using gsub, to make it more explicit that these variables corresponded to the time and frequency decompositions.

Part 5 code

Using the dplyr package, the mean of the columns is calculated very easily. The groupby function selects how the data is to be combined for the calculation in the subsequent summariseall step. In this case the data is to be averaged across participant ID and physical activity. The summariseall step then calculates the mean values for all of the columns. Finally arrange organizes the resulting data in ascending order based on participant ID followed by physical activity. The result from this is stored in the Avg_Data variable which is then written to a text document called Tidy_data.txt participant ID and physical activity.