Senior Research

README for Rethinking the Application of Technology in Hypertension Control Overview The .Rmd files are split up by specific tasks and should be run in a particular order. The “data_assembly.Rmd” should be run first. It loads the data included in the NHANES_DATA folder (included with these files) into R-studio and manipulates the data into a usable format. Therefore, it is important to change the working directory for this file to match the location of the NHANES_DATA folder. The data in the NHANES_DATA folder was downloaded from the NHANES website (CDC, 2019). The “clustering.Rmd” should be run second. It clusters the data for the “lin_cluster.Rmd”, which builds linear models on the clusters. The “linear_model.Rmd” should be run third. It builds the baseline model, and also crucially creates the data frame “mse”, which is used to store the models’ performance for the rest of the study. “Lin_cluster.Rmd”, “decision_trees.Rmd”, and “random_forest.Rmd” all build different types of models and records their performance in the “mse” data frame. They can be run in any order. The “comparison.Rmd” compares the performance of the models built in “linear_model.Rmd”, “Lin_cluster.Rmd”, “decision_trees.Rmd”, and “random_forest.Rmd”, and thus should be run after all these files. The “maha_comp.Rmd” finds a cut-off value for the mahalanobis distances. It can be run as long as the “random_forest.Rmd” has been run. Lastly, “misc.Rmd” simply contains code that was developed for the study, but went unused. See the order listed on page two. Please find the App for the research at this link: https://buehlere.shinyapps.io/App-1/