View profile

Micromorts | Next - Issue #36

Hi there,
Last week I came across this beautiful New York Times visualisation, which answers a simple question: Are your neighbours mostly democrats or republicans? The tool illustrates how much we can know by looking at the same mundane datasets differently.
Anyway, let’s dive in.

Five Stories
A peril of statistics is that it makes probabilistic estimates sound deterministic. How do we communicate uncertainty? Put in other words, how do we convince ourselves that our results are “true” and not a fluke?
Effect size is one such measure. They are general rules of thumb that guide our understanding of essential metrics. This vignette explains different rules of thumb with examples from effectsize package in R.
Clinical trial data is messy to handle. tidyCDISC is a Shiny app that can create many tables and figures based on standard medicine requirements. The app provides extensive capabilities to explore data over the population and for a given patient.
High-level features of the app allow users to produce customized tables using a point-and-click interface, examine trends in patient populations with dynamic figures, and supply visualizations that narrow in on a single patient profile.
Watch the YouTube talk or read the vignette to learn more (recorded at R/Medicine 2020).
Sir David Spiegelhalter is a British statistician who, between 2007 and 2018, was the Professor of the Public Understanding of Risk at Cambridge University (what a cool title!). In this creative video, he discusses how little we understand about general risks.
Fun fact: When our “prize” is death, and its probability is tiny, a different scale was proposed in 1979 by R.A. Howard to convey the chance of dying. A micromort (MM) is a one-in-a-million probability of death.
Prof William Deming is known for his contributions to quality improvements, among numerous other things. Japanese companies like Toyota and Mazda adopted his science and production principles, making them world-famous as we know them today.
This is a free 12-day course to understand Deming’s philosophy and quality control statistics. I haven’t read all of it but this story about school buses and the weather was superb.
Version control for your codes is a good idea — everyone knows that. Have you thought of version control for your writing as well? Rose brings out several points to convince us: when we need version control.
  • I need to re-work this presentation for a new audience,
  • I’ve spent the last month revising a draft according to my advisor’s feedback, and now she’s saying we should undo all of that and go back to the original framing!
  • I spilt orange juice on my laptop and lost all my files.
Anyway, long answer short: create a Github account and put your writing there. It doesn’t have to be public, you can totally do a private repository.
Four Packages
Ever looked at a plot and wanted to get the data from it? juicr provides a GUI interface to tools for extracting data from scientific images like scatter or bar plots. See the vignette here.
effectsize provides utilities to work with indices of effect size and standardized parameters, allowing computation and conversion of indices such as Cohen’s dr, odds ratios, etc. See the vignette here.
groundhog allows you to fix your package version to a specific day. R scripts that rely on packages are reproducible by ensuring that every time a given script is run, the same version of the used packages is loaded. It works by simply replacing library("dplyr") with groundhog.library("dplyr", "2022-06-22"). See the vignette here.
reportfactory facilitates workflows with multiple R Markdown documents. They can execute them in parallel and store outputs in well-organised folders. Check the vignette to learn more!
Three Jargons
Git is a version control system that allows tracking changes in any set of files. They’re useful for coordination between programmers working on the same code base. Learn more.
The control chart is used to study how a process changes with time. It has a central line for the average, an upper line for the upper control limit, and a lower line for the lower control limit. These lines are determined from historical data. Learn more.
Some probabilities are small but significant, like death by car accidents. For such rare occasions, we use a special metric. A micromort (MM) is a one-in-a-million probability of death. Learn more.
Two Tweets
Arthur Welle
Pipes in #RStats! Functions are verbs, aplied to objects. With pipes "|>" we arrange functions in the order that we would think of the actions. Readability for the win! (And I like to read the pipe as "and then...")🎂🥣🥮👩‍🍳🔪🍴
Spencer Schien
Saw #JurassicWorldDominion and thought it was bad...real bad. I wondered what others thought, so I scraped reviews from IMDb. Here's a repo with #Python scraping code, #RStats code to process and make this word cloud, and of course, the data:
One Meme
That's a wrap!
I hope you learned something new today. As always, your feedback and suggestions are always welcome. See you next week!
Did you enjoy this issue? Yes No
Harshvardhan @harshbutjust

A short and sweet curated collection of R-related works. Five stories. Four packages. Three jargons. Two tweets. One Meme.

List of all packages covered in past issues:

In order to unsubscribe, click here.
If you were forwarded this newsletter and you like it, you can subscribe here.
Created with Revue by Twitter.
Knoxville, Tennessee.