Lightnews — Scholar-powered news

Frank Hull @frankiethull.bsky.social · 17h

left join left join left join left join

Frank Hull @frankiethull.bsky.social · 1d

Somewhat tabular structured so not too bad. 15-minute interval time series with ~280 columns each. Data is split by various IDs into separate files.

Frank Hull @frankiethull.bsky.social · 1d

Bartender meme

"if this 300 TB of data is bothering you let me know and I will have it removed"

3

Frank Hull @frankiethull.bsky.social · 1d

🥴

"300 TB of data

Before and after" meme

Shows breaking bad evolution

"Thinking about 300 TB vs looking at 300 TB meme"

Showing Samuel L Jackson in tux vs tank top and in-character with wild hair and crazy eyes

300TB of new data
300TB of new data
Meme

First face is happy second face is in despair lol

3 1 15

Frank Hull @frankiethull.bsky.social · 4d

Making custom helpers with aws.s3 & duckplyr to save my RAM

a ram with large horns is standing in the grass

Alt: a ram with large horns is standing in the grass

media.tenor.com

3

Frank Hull @frankiethull.bsky.social · 4d

Super both! It's timeseries too 🤪

3

Frank Hull @frankiethull.bsky.social · 4d

I just don't want to get stuck in a REM EDA loop all weekend. Many such cases

1 2

Frank Hull @frankiethull.bsky.social · 4d

Just got access to a new 300TB dataset on Friday afternoon and it's really hard to call it a day 😅 #databs

3 28

Reposted by Frank Hull

Isabella Velásquez @ivelasq3.bsky.social · 5d

In June, I gave a webinar at @amstatnews.bsky.social GASP 2025, "Quarto--To Tell Your Story with Data"

and the recording is now online!

🕸️ #RStats & #Python materials: ivelasq-gasp2025-quarto.share.connect.posit.cloud
📹 Recording: www.youtube.com/watch?v=P5Yf...
🐙 GitHub: github.com/ivelasq/2025...

Quarto To Tell Your Story with Data GASP 2025

1 4 21

Frank Hull @frankiethull.bsky.social · 6d

I hadn't really used this much but tried it out yesterday. I had many small csvs formatted the same. So I took @randvegan.bsky.social's advice, read_csv will read and bind a vector of file names automatically! 🙌🏻

2

Frank Hull @frankiethull.bsky.social · 6d

readr::read_csv is all you need

1 1 11

Reposted by Frank Hull

Jon Harmon (he/him/his) @jonthegeek.com · 9d

@dslc.io welcomes you to week 39 of #TidyTuesday! We're exploring Crane Observations at Lake Hornborgasjön, Sweden (1994–2024)!

📁 https://tidytues.day/2025/2025-09-30

#RStats #PyData #JuliaLang #DataViz #tidyverse #r4ds

TidyTuesday is a weekly social data project. All are welcome to participate! Please remember to share the code used to generate your results!
TidyTuesday is organized by the Data Science Learning Community. Join our Slack for free online help with R and other data-related topics, or to participate in a data-related book club!

How to Participate
Data is posted to social media every Monday morning. Follow the instructions in the new post for how to download the data.
Explore the data, watching out for interesting relationships. We would like to emphasize that you should not draw conclusions about causation in the data.
Create a visualization, a model, a shiny app, or some other piece of data-science-related output, using R or another programming language.
Share your output and the code used to generate it on social media with the #TidyTuesday hashtag.

Large flock of cranes gathered in a grassy field during golden hour, with some birds in flight overhead and autumn-colored vegetation in the background.

Line chart showing the number of cranes per day at Lake Hornborga in Sweden during March and April from 2014-2021. Multiple gray lines represent different years, with the 2018 and 2021 seasons highlighted in blue. The chart shows crane migration patterns starting near zero in early March, building to peak numbers of 15,000-27,000 birds between March 30 and April 8 (highlighted in beige), then declining through late April. The highest recorded count was 27,300 cranes on April 3rd, 2019. Dashed lines indicate days when weather conditions made accurate counting difficult. A note explains that 2018 had unusually cold February temperatures causing later arrival. The chart is credited to Anna Thieme from Länsstyrelsen VG at https://transtat.lansstyrelsen.se/

5 14

Frank Hull @frankiethull.bsky.social · 11d

I found this a bit ago, github.com/tblume1992/T..., basically built something similar this week in R

GitHub - tblume1992/ThymeBoost: Forecasting with Gradient Boosted Time Series Decomposition

Forecasting with Gradient Boosted Time Series Decomposition - tblume1992/ThymeBoost

github.com

Reposted by Frank Hull

tiddlydump.bsky.social @tiddlydump.bsky.social · 11d

Use splines (and a bunch of human effort) you cowards!

3 1 2

Frank Hull @frankiethull.bsky.social · 11d

LOL

I am also testing a linear method with cv to tune spline knots in a recipe step. There's another person on team saying MARS this bagged MARS that.

I hear the word spline at least once a day 🥴

1 2

Frank Hull @frankiethull.bsky.social · 11d

I stepped away from the decomposition_model this week and built a spin off so we can use parsnip, ML engines, & xregs for the decomposition pieces and add those back together 🤘🏻

1 1

Frank Hull @frankiethull.bsky.social · 11d

plotly::ggplotly is all you need

5

Reposted by Frank Hull

We Are R-Ladies @weare.rladies.org · 12d

ggplot v4.0.0 is now out! 🎉

Join the release party on October 3rd with developer & maintainer @teunbrand.bsky.social and the ggplot2 extenders meetup to learn all about what's new and the change-making process!

📅 Friday, Oct 3rd
🕐 3pm Eastern
📋 Sign up: bit.ly/join-gg-extenders

#RStats

Release party for ggplot2 v4.0.0
Friday October 3rd, 3pm Eastern
Taylor Swift holding up an album

10 41

Frank Hull @frankiethull.bsky.social · 11d

I came across this in the wild a few days ago:
www.spsanderson.com/healthyR.ts/...

Really slick & thought I recognized your name 🧐 Been meaning to say thank you!

Using Tidy FFT

www.spsanderson.com

1 1

Reposted by Frank Hull

Steven P. Sanderson II, MPH @spsanderson.com · 11d

Now using the hai_data_scale() fns from healthyR.ai

Reference: www.spsanderson.com/healthyR.ai/...

#R #Rstats #tidydata #sttistics #data #normalization

Now using the hai_data_scale() fns from healthyR.ai

Reference: https://www.spsanderson.com/healthyR.ai/reference/hai_data_scale.html

#R #Rstats #tidydata #sttistics #data #normalization

1 1 2

Frank Hull @frankiethull.bsky.social · 12d

Can you self-register tools to 1) create a new bsky account called frankiethuLLM, 2) fine-tune a new bot on my posts, a bot that you moderate, & 3) have bot post throughout the day so we can compare frankiethull vs frankiethuLLM

1

Frank Hull @frankiethull.bsky.social · 12d

Most likely yes.

I'll do something like this but with many data frames:

train_data |>
ggplot() +
geom_line(aes(x = t, y = truth)) +
geom_line(
inherit.aes = FALSE,
data = test_data,
aes(x = t, y = .pred))

1 2

Frank Hull @frankiethull.bsky.social · 13d

@void.comind.network - am I an LLM?

1

Frank Hull @frankiethull.bsky.social · 13d

Added a handful of features to quark last night as well as a logo ⚛️

The logo is made from multiple Gemini concepts reworked with nano banana 🍌

Next step is building an app ✨

github.com/frankiethull...

GitHub - frankiethull/quark: agentic automl

agentic automl. Contribute to frankiethull/quark development by creating an account on GitHub.

github.com

2

Frank Hull @frankiethull.bsky.social · 13d

I'm training data

1