LiveThread of #cascadiaRConf afternoon session [with apologies to @tladeras for missing the first lightning talk; my dog needed food]. Now building a college basketball simulation by Kevin Floyd. 
@reikookamoto1 lightning talk about bringing analysts and technical writers together through internal packages to support teams.
#cascadiaRConf
#cascadiaRConf
A theme of many of the talks is the need for development to market/show how the packages solve key problems.
Keisha R. Harrison on Defining Complex Microbial Systems within Kombucha Fermentations Using METACODER and PYHLOSEQ Libraries
or what's the biology of kombucha...
#cascadiaRConf
or what's the biology of kombucha...
#cascadiaRConf
How cool...
Erin Dahl talks about microshades for color accessibility in complex data.
Types of color-vision deficiency (CVD)
Existing solutions have few colors. {microshades} expands from 8 colors to 30
An example: wow!
Jill Levine talking about the Clio package for clean historical data.
@jdlevine presents a visual derived from the data [clio.infra] on book titles
Next up, Dror Berel: Been there, done that. Practices I adopted from R that will guide me through learning additional programming languages
The path:
Augment your wheel, don't reinvent it.
@Edgar_Zamora_ talks about redesigning workflows
Overview:
It is a worthwhile investment:
Next is Asha Yadav: Data-Informed Decision Making in Part C Early Intervention using Shiny R
Evaluate an intervention and make the evaluation accessible.
Impact...
User-centered tools are key
Shiny empowers the users.
Accessibility is not just about colors and the like, familiarity too....
Howard Baek ( @howard_baek ) : Feedback Report: A Web-Based Shiny Dashboard for displaying Patient-Reported Outcome Measures for Patients in Addiction Treatment
A website in 5 minutes! Sean Kross ( @seankross ): Create a Personal Website in 5 Minutes with Postcards
Get an easy win, fill in that top left hand corner with a personal website built in *postcards*
The final lightning talk: Katie Jolly ( @katiejolly6 ): Designing graphics to post online: What I’ve learned from the (sometimes helpful) comments on my maps from Twitter and Reddit
Alternate title:
No truer statement has been written:
Exactly. What I think a graphic means applies, perfectly, TO ME!
Learn from coders by sharing code:
The lightning talk format and all the lightning talks were excellent. You really otter be at #cascadiaRConf
Using R Session begins with a talk from @JohannWindt The PlayerMakeR Pipeline - Using R across the data pipeline with the Vancouver Whitecaps Football Club, an example using wearable technology.
The wearable tech illustrated:
@JohannWindt presents an absolutely awesome schema for evaluating the cost-benefit of new instruments and sources of data
Esther Needham and Bryan Blanc: Using Shiny Dashboards to Understand Bus Transit Delay and Sketch Solutions for King County Metro Transit
Bus usage increased (as a % of transit ridership) under COVID...
An amazing shiny:
Evaluating interventions in a shiny.
Cordero Ortiz: PurpleAir PM2.5 Modeling in Portland, Oregon. Missed the title slide but here is what PM2.5 is.
The sensors
The sensor array in Portland
There's an R package for the sensors. Who knew?
Time series data too
There's a ton of data with a ton of layering! For R, no problem.....
A random forest to generate a predictive model.
All that data, all those models, shiny brings it together.
The Q&A with a focus on how to get the users to understand what is provided and for adapting or baselining to help them with data understanding.
The final session is Education & Community. Leading off is Kristin Bott talking about how they built community at Reed. Faculty have heterogeneous abilities and needs, students have more of both.
@RhoBott describes the team with a key need for emotional intelligence because it meets the needs where those needs are and building resources to enhance availability to encourage and empower.
People first, code second. A wonderful mantra.
Clear scoping is vital.
Waffle emojis
Njesa Totty: Validating the assumptions of bootstrap intervals for responsible implementation using the R package bootEd
Helping students to understand the underlying assumptions is key. So simulate it!
Nick Paterno: Teaching Non-STEM Students to Code will conclude the Education & Community session
Small colleges/universities covering a huge range of base conditions:
The talk appears to deploy Python indexing
Recounting the ubiquitous experience of intro stats instructors using R: the intersection of coding and stats and students not entirely connecting why each is non-ignorable.
Use `pkgdown` to minimize troubleshooting.
For many instructors, learning the language is important. That's true. But for students, using R as a tool to advance their needs is where the early value lies. Understanding the power of the tool can motivate that final step to the language as having intrinsic value.
Q&A
Moving to the concluding remarks, @ellis_hughes
Things Ellis learned
I cannot echo this enough! Thanks to all those whose hard work yielded an excellent conference!
Already looking forward to it.
@threadreaderapp unroll
• • •
Missing some Tweet in this thread? You can try to
force a refresh
