In this working group we learn about methods and ideas for confronting models with data, and about data science generally. Anyone is encouraged and welcome to attend.
| Info |
|---|
We have concluded our working group meetings for spring 2022. Perhaps we will see you again in Fall 2022. |
Meeting schedule and topics:
Zoom link – For the spring 2022 semester, we
...
met at 11 a.m. Wednesdays (Mountain Time) on Zoom.
...
Meeting schedule and topics:
Spring 2022
– building an R package (for CRAN, GitHub, or Bioconductor).
how to make a minimal R package in the style expected by CRAN. We might want to work through https://r-pkgs.org/whole-game.html from https://r-pkgs.org (a book focused on this topic).
Here is a quick tutorial, that is similar to the “whole game” tutorial linked above, but with a few twists.
hosting options and considerations
– In advance of our meeting, please watch https://youtu.be/dag9l0GFci8 from the 2021 useR conference and come ready to discuss. This 40 minute talk (plus ~15 minutes of Q&A) covers a diversity of topics and provides some food for thought. See The United States Research Software Engineer Association and their job board.
– Damir Pulatov – How to write unit tests for your R code why you should do it.
Brief shoutout - useR is an R conference and there is a call to offer tutorials (link goes to the useR page) at the conference. Proposals are due by Feb 2015. Might be a way to ‘level up’ for some of us. Suggested topics are broad and introductory (e.g., git with R, data visualization basics, etc.)
– … R, Iteration in R – loops, apply, map functions, Functional(s) and programming in R Eryn McFarlane - back to very basics. link here
– No meeting this week. Consider watching this recorded talk instead: https://www.youtube.com/watch?v=B7TBfJrofQM
Not related, but I found this article on the use of AI insightful – https://www.technologyreview.com/2021/07/30/1030329/machine-learning-ai-failed-covid-hospital-diagnosis-pandemic/
– visit with Nick Anderson, security engineer at Facebook, undergrad degree in math from UW, Wyoming native. We will learn about his education and career path and ask him questions like (please bring your own too):
what advice would you give to yourself as a finishing undergrad at UW, what do you wish you had known at that point?
what things would you encourage students and postdocs to do to prepare themselves for a career in security, or more broadly a career in a technology company?
what are some of the main areas in which technology companies deal with security?
what is your area of speciality within security and how has this changed over time?
what are areas that you are excited learning more about in security?
in what ways do machine learning experts and data scientists contribute to security teams?
– Development environments (a.k.a. pimp my rIDE)
Discussion of last week’s meeting with Nick Anderson.
Code editors – Share, demo, try out different IDEs / editors: Visual Studio Code, BBedit, emacs ( read and write files remotely with built-in ssh support: /ssh:teton:tmp.txt; split screens and multiple windows on same buffer), vi, Rstudio, atom, Eclipse.
– Some learning about Apache Arrow – 🌟 in advance, please watch short talk by Wes McKinney (25 min, demo starts around 18 min) and look over this short introduction.
we will discuss and share thoughts on the video and short introduction
We can experiment with these simple examples. https://arrow.apache.org/docs/r/
We will also step through this vignette, which is part of the R package ‘arrow’. It calls for working with 37 GB of data from AWS S3. Consequently, I suggest we skip trying to run the code ourselves, but instead we step through the vignette and work to understand what it is doing without executing it ourselves.
Extras:
– 🌴 spring break 🏖 – no meeting ⛷
– Interactive visualizations with Shiny – demo and hands-on – Heili Lowman – here is the shiny app that Heili Lowman built in the demo.
For inspiration when building your own Shiny applications, there are lots of examples online in the Shiny Gallery. There are also a number of examples of Shiny Dashboards, if you prefer that layout.
Additional tutorials are available here and here, and lots more details are available in the textbook Mastering Shiny by Hadley Wickham.
The Shiny Widget Gallery provides all the code necessary to add any kind of widget you like. The ‘shinyWidgets’ package also has more customizable widgets if you would like to offer the end user more freedom in their selections.
Today’s lesson was adapted from a lesson created by Allison Horst for the R Ladies Santa Barbara chapter (GitHub repo for the lesson here). A copy of today’s script ('app.R') can be found below.
Alex's example with >12,000 possible plots that are calculated on the fly. This was my first Shiny app, which I use for teaching.
and remainder of spring semester – we are going to put the working group on hold for the rest of the semester and will see about regrouping again in fall 2022. Thanks everyone for participating and sharing your interest and enthusiasm for learning things together. Don’t forget the lesson about getting accustomed to failing. Get out there and push yourself and see what you can do. And have fun doing it!
...