ID529 Course Recap

moral

“I will let the data speak for itself when it cleans itself.”
—Allison Reichel



You should be incredibly proud of yourselves not just for getting to the finish line of ID529, but for all the work, the trial and error, the questions and the fantastic attitudes you’ve shared with all of us.
—Christian, Hodu Dad

key principles of ID529

  • understand your data, goals, and conceptual frameworks
  • organize your work
  • make your work more pleasant for yourself
  • make your work navigable for others
  • get help from the community
  • reduce tedium and manual processes
  • make collaboration easier
  • make communication easier

tidy workflows

the workflow in tidyverse

data analysis work can’t be just modeling

remember, we can't just focus on modeling

practice

every aspect of our analysis workflows will benefit from practice

use R projects and packages to stay organized

  • have a roadmap of what you’re aiming to accomplish
  • have a changelog or timeline or anthology of decisions made on your project
screenshot of new project creator

use git to avoid

  • going on archaeological digs on old email threads
  • uncertainty about which version of the code is the most up-to-date
  • uncertainty on what’s the most recent data
  • uncertainty on if issues have been solved
  • zip files of file salad
use git to avoid v_final

make your work more pleasant

using modern tooling in R and on GitHub

  • you can have joyful code instead of dreadful code
  • you can feel confident about your code instead of leery and uncertain
  • your projects can be colorful and expressive, not bland and joyless
this one brings joy, marie kondo

making your work navigable for others

sharing your work through markdown files on GitHub is great, but you can also

  • blog about your work
  • tweet / post about it on social media
  • and make interactive web pages to play with the data / findings

get help from the community

connect with the RStats community to learn more happy-making practices

reduce tedium and manual processes

  • sometimes tedium is necessary; however:
  • often solutions already exist to do what you’re trying to do!
  • or you can be clever about restructuring what you need to do to leverage tooling that already exists

make collaboration easier

  • use git & github
  • use clean & considerate code
  • make your work easily reproducible

a screenshot of people I follow a screenshot of people I follow

make communciation easier

  • no one has every said “it’s a shame we have all this great documentation”
  • reproducible examples will make getting help on your challenges Inf% easier
  • use R Markdown and literate programming practices to make your analysis transparent and engaging for your colleagues and collaborators.
  • a fantastic ggplot can do more than speak 1000 words: it can tell a good story with a coherent narrative and resultantly capture and intrigue viewers.

examples of rmarkdown documents example of rmarkdown document with figure

resources