@DataAngler@vis.social
@DataAngler@vis.social avatar

DataAngler

@DataAngler@vis.social

I am a data scientist who programs in R and SAS. Don't ask me to pick one. I am also particularly interested in equity in education as well as gender equity and equity in policing. Occasionally fly fishing. My posts are my own.

This profile is from a federated server and may be incomplete. Browse more on the original instance.

pluralistic, to poop
@pluralistic@mamot.fr avatar

"who knew poop shaming was even a thing?"

https://www.reddit.com/r/LateStageCapitalism/comments/194trv8/who_knew_poop_shaming_was_even_a_thing/

#DisciplinaryTechnology #Bossware #SIgns #Poop

DataAngler,
@DataAngler@vis.social avatar

@qkslvrwolf @pluralistic Hoping that the mean is from a severely right-skewed distribution...

davidbisset, to random
@davidbisset@phpc.social avatar

Friend: "What does your downloads folder on your computer look like?"

Me: 🤔

DataAngler,
@DataAngler@vis.social avatar

@headrift @davidbisset select *

teunbrand, to random

I'm excited for the new ggplot2 version coming up that I helped out with. So I thought a fun game might be to post a plot every day and people can guess what the new thing is. No cheating by looking at the NEWS.md file!

Let's start with an easy one

DataAngler,
@DataAngler@vis.social avatar

@teunbrand Annotations behind data points?

DataAngler, to random
@DataAngler@vis.social avatar

Just learned that I needed to go back a version in to keep using the groundhog 📦 and load versions of packages that I was using. Future me better appreciate this...

DataAngler,
@DataAngler@vis.social avatar

@odr_k4tana @ingorohlfing That's a life goal. Containers.

DataAngler,
@DataAngler@vis.social avatar

@ingorohlfing Thanks for the tip. I do like how groundhog 📦 installs versions from a snapshot in time so that I don't have to specifiy which versions I want, but it works less seamlessly when R itself is updated.

defuneste, to random
@defuneste@fosstodon.org avatar

I am going to say it but {here} is great if you are working in a project oriented workflow (yes that is the first line of its description). If you are using it in an interactive session, calling it a bunch of time in "different places" (in code and file structure) it will not help!

Please use a project oriented workflow! It will not just help your coworkers it will also help you avoiding some mistakes.

DataAngler,
@DataAngler@vis.social avatar

@defuneste I recently just handed an RStudio project to another user (like put it in a whole new directory) and said when you call here(), all of these links within the project will just work. I haven't heard from them, so I'm trusting it worked as planned. 🙂

DataAngler,
@DataAngler@vis.social avatar

@defuneste I'm here for {here}. I guess I cannot think of a case in which I would call it multiple times in a project unless my stuff was all over the place...

lf_araujo, to random
@lf_araujo@mastodon.social avatar

Is there a dashboard where I can follow download statistics (with visualizations) for certain CRAN packages?

DataAngler,
@DataAngler@vis.social avatar

@aghaynes @lf_araujo There's also this {flexdashboard} example: https://testing-apps.shinyapps.io/flexdashboard-shiny-crandash/#section-dashboard

joranelias, to random
@joranelias@mastodon.social avatar

I know it violates the data component/non-data component dichotomy but I wish it were easier to map aesthetics to categorical axis labels

DataAngler,
@DataAngler@vis.social avatar

@joranelias I, too, default to using scale_fill_manual() and had a maddening time mapping my colors onto factor levels with alluvial diagrams.

kristinHenry, to random
@kristinHenry@vis.social avatar

Hey vis.social folks!

Why do we try to maintain our Local Timeline as on-topic?

There are so many reasons. But one big one is that our Local Timeline is what other folks see when they are checking out if they'd like to join our community or not.

We are a topic-based instance, so when our Local Timeline is all over the place and there's no Data Viz or Art or any Creative process...What are potential community member supposed to make of this?

--vis.social moderation team

DataAngler,
@DataAngler@vis.social avatar

@kristinHenry I think it means that we are a community who works in data-visualization related fields or who has an interest in data visualization (or artistic productions in various visual mediums). But we also have other interests. I've often wondered if maybe this shouldn't be my instance due to the fact that I don't post on exclusively...

tanya_shapiro, to random

data friends, what are some things you want to learn this year?

i have a CVS-receipt-long list of things... but maybe top of the list i'd say: webR/wasm, more js & d3, svelte, docker, more on app dev (hosting, security, etc).

DataAngler,
@DataAngler@vis.social avatar

@tanya_shapiro I want to do more analyses in a Bayesian framework, figure out Quarto in , , and get into double digits in Advent of Code.

Uriel, to random

Are you switching from RMarkdown to Quarto? If so, why?

It seems super "in" with the community, but in my case, I'm not sure it's worth the extra coding. For context, I'd be using it to make manuscripts.

DataAngler,
@DataAngler@vis.social avatar

@Uriel I used in recently to make a set of parameterized reports, but I found it very difficult to render to MS Word and could not control spacing in the reports w/o strange hacks. {gt} tables were rendered, also, with inconsistent formatting that was difficult to standardize. Next time, I might see if is easier.

KidsData, to ukteachers

Among California kindergartners, chronic absenteeism has soared—from 16% pre-pandemic to 40% in 2021-22 and 36% in 2022–23.

More findings in this important new Policy Analysis for California Education report by Hedy Chang, Belen Chavez, and Heather Hough.
➡️ https://bit.ly/3HEyvY

@edutooters @educationecon

DataAngler,
@DataAngler@vis.social avatar

@KidsData @edutooters @educationecon broken link?

o_simardcasanova, to random
@o_simardcasanova@mastodon.social avatar

Mastodon / hive mind

Does anybody have a simple solution to add Hawaii and Alaska in a box in a map?

I tried with this one, but it was VERY complicated

DataAngler,
@DataAngler@vis.social avatar

@o_simardcasanova The {usmap} 📦 is an option.

library(tidyverse)
library(usmap)
ggplot(usmap::us_map()) +
geom_sf() +
coord_sf() +
theme_void()

Bikeyface, to random
@Bikeyface@better.boston avatar

I thought the surprise $5000+ bill from the IRS was the worst part of the day… but then I was hit but a driver. Left hook, I had a green. He clipped the back of my bike.

(I’m okay, bike is okay, I’m just shaken. Yes I filed a police report, because the driver took off and I wanted it documented)

DataAngler,
@DataAngler@vis.social avatar

@Bikeyface @glightly yikes. Glad you are okay. I wish cycling would not feel like a roll of the dice

cct_datascience, to foss

Inspired by a recent collaboration, we’re curious about approaches for publishing a research codebase for others to reuse and cite…beyond releasing an R package! 🤔 :rstats: 📦

folks - are there circumstances where you’d publish code without structuring it as an R package? Are there examples that spring to mind?

DataAngler,
@DataAngler@vis.social avatar

@cct_datascience I think it is common for people to publish their code in a repository like OSF so that others can re-analyze study data. Many researchers write R code but fewert build R packages.

DataAngler, to cycling
@DataAngler@vis.social avatar

Going home.

DataAngler, to Birding
@DataAngler@vis.social avatar

Highlights of in Ashland Nature Preserve in :

Two downy woodpeckers flushing insects (?) in the brambles
Red shouldered hawk alone in a lofty pine tree
Pretty sure I saw a hairy woodpecker as it was larger than a typical downy

And app from indicated it heard a golden-crowned kinglet but I did not get it

DataAngler, to random
@DataAngler@vis.social avatar

One side effect of banging your head against the wall with #adventofcode in #rstats is that when your job needs you to write a function to run 100 times on a data set with reproducible seed and store each run in a list, you're like, hey no problem.

DataAngler, to Birding
@DataAngler@vis.social avatar

After a steady rain, the tops of trees are dotted with Turkey vultures airing out their wings.

DataAngler, to datascience
@DataAngler@vis.social avatar

Having a good colleague is knowing just which letter to change in several variables in their code to be able to render a whole new set of .

DataAngler, to Birding
@DataAngler@vis.social avatar

Saw a black-capped chickadee and a tufted titmouse in a hedge. Black-capped chickadee if’s via sound id in the Merlin app.

DataAngler, to StarTrek
@DataAngler@vis.social avatar

I found out that my boss has no knowledge when he did not share in my freakout after I said 7 of 9 variables did not have balance in the propensity scores.

DataAngler, to random
@DataAngler@vis.social avatar

A non-trivial part of doing Bayesian data analysis is to get the darn computer to work. I cannot get to call Stan due to a well-known issue with managed devices.

DataAngler, to random
@DataAngler@vis.social avatar

I think I learned today that your parameters in should not have the same name as variables in data frames. I could not render parameterized reports correctly. E.g.,. if my variable in my data is named "school" I should use something else in the YAML header like

params:  
 org:
  • All
  • Subscribed
  • Moderated
  • Favorites
  • •
  • JUstTest
  • InstantRegret
  • mdbf
  • osvaldo12
  • magazineikmin
  • GTA5RPClips
  • rosin
  • thenastyranch
  • Youngstown
  • cubers
  • slotface
  • khanakhh
  • kavyap
  • DreamBathrooms
  • anitta
  • Durango
  • everett
  • ethstaker
  • cisconetworking
  • Leos
  • provamag3
  • modclub
  • ngwrru68w68
  • tacticalgear
  • tester
  • megavids
  • normalnudes
  • lostlight
  • All magazines
    •