Survival package for R

Summer Research, 2018

Students: Ashley Jacobson and Victor Wilson

Faculty advisor: Shannon Pileggi

Objective

The objective of this summer research is to create an R package that can execute parametric and non-parametric survival analysis techniques similar to those in Minitab.

Deliverables

A GitHub repository which contains the following:

1. An R package for the survival functions.

2. A log of hours spent on the summer research by each student, which includes date, hours, and activity summary.

3. A presentation to the Statistics Department.

4. A presentation at the CSM annual research conference.

5. A manuscript to submit to the R Journal detailing the work and submit an abstract to the RStudio Conference.

Specific Aims

1. Utilize GitHub to collaborate on project materials and updates.

• Karl Broman's github tutorial

• Jenny Bryan's Happy git with R.

• DataCamp's Introduction to git for data science course (good for learning command line, not necessary if using the RStudio IDE).

• Also check out using version control with RStudio and this video on Git and RStudio.

2. Adhere to good programming practices.

• Write all R code according to Hadley Wickam's Style Guide.

• Use the tidyverse style guide for an additional reference.

• Learn about how to write R functions from DataCamp's Writing functions in R course.

• Use Hadley Wickham's R for Data Science book as a reference (Ch19 also discusses functions).

  3. Create an R package that contains survival functions. At a minimum, this should be downloadable through devtools; as time allows, consider putting it on CRAN.

• DataCamp's blog post R Packages: A Beginner's Guide

• Hillary Parker's blog post Writing an R package from scratch

• Hadley Wickham's R pacakges book.

• RStudio's video on Package writing in RStudio.

  4. Provide documentation for the R package.

• Use the roxygen package to document code.

• Write a vignette to accompany the package.

• Consider using pkgdown to create a website.

  5. Review existing R packages for survival analysis.

• A comprehensive list: CRAN Task View: Survival Analysis.

• Pay attention to: survival, fitdistrplus, flexsurv, and survminer.

  6. Create functions for items that are not currently easy to achieve in R. (Make sure that these cannot be accomplished in the existing R packages). Pay attention to the parameterization of a distribution, which is often different between Minitab and R.

Parametric:

• Lab 2:

  • Compute survival probabilities based on a specified parametric distribution.
  • Display an estimated survival, hazard, or cumulative hazard curve based on a specified parametric distribution (singular as well as by groups).
  • Estimate median survival time based on a specified parametric distribution.

• Lab 3:

  • Estimate parameters based on a specified parametric distribution.
  • Compute the Anderson-Darling test statistic, or otherwise appropriate measure of fit, for a specified parametric distribution.
  • Estimate various statistics, including median, mean, standard deviation, and percentiles of survival time based on a specified parametric distribution.
  • Produce a qqplot to examine distribution fit.

Non-parametric:

• Lab 4:

  • Plot estimated hazard/cumulative hazard functions

• Lab 5:

  • Compute the Wilcoxon test statistic and p-value to compare survival experiences of two groups.

7. Consider also converting some of the functions to a Shiny App.

• Shiny tutorial on RStudio

• DataCamp's Building Web Applications in R with Shiny

8. Other potentially useful DataCamp courses.

• Working with the RStudio IDE Part 1

• Working with the RStudio IDE Part 2

• Introduction to the tidyverse

• Data visualization with ggplot, 1

• Data visualization with ggplot, 2

• Data manipulation in R with dplyr