Frank Hull

Automating Repetitive Reports with Quarto, 2025 MAPOR Fall Webinar Series.

Next week! I am giving a webinar on "Automating Repetitive Reports with Quarto" as part of the 2025 MAPOR Fall Webinar Series!

Save serious time and eliminate errors by letting Quarto do the heavy lifting 🏋️

I would love to see you there 💚💛💙
www.mapor.org/2025-mapor-f...

Mostly learned from this book & looking at open source packages on GitHub for ideas r-pkgs.org

For any extension package related to tidymodels, tidymodels, etc. they also contain vignettes for developers that come in handy

You #rstats, all I'm going to say is that I'm so excited for the amazing speakers and topics we have for you on the R+AI lineup. Official marketing posts are coming but I am so excited, it's hard to not talk about it.

Thanks to the @posit.co team for adding the @bsky.app icon to the top right of the navbar at quarto.org.

With the latest @quarto.org update, the 🦋 Bootstrap icon works on your personal Quarto homepage too 🥳
#rstats

Relatable. Do everything in UTC before converting to local time (& boycott DST) lol

A very important detail from a data sage 🧙🏻🙌🏻

But eventually need to use that data somewhere else. Makes sense to start in a project for better organization. But if pieces can be helpful downstream for team, just make a package. Now we have 30+ or so modularized data, modeling, visualizing, branding, reporting packages that work together 🙌🏻

I'm skipping over a lot. Originally I would get stuck thinking "should this be a script or a notebook" eventually I'd have to pull the code from a notebook and place in a script. Modularizing scripts like 00_get_data, 01_clean_data, 02_feature_data, 03_regression, etc.

I haven't mentioned it (until now) but "all you need" is a survey analysis I'm conducting. Eventually scrape my profile with bskyr and see which "all you need" is ranked highest and truly all you need.

6. Everything should be a project or package

I'm almost to the point where everything should be a PACKAGE. This has saved me so much time this year.

7. Vignettes, NEWS, Issues

Helpful for the developer, user, and managers to follow along. Or even your future self 🤗

4. Everything will be used again

This is the opposite of my early script days. It's the mindset to build modular. Eventually I became modular with scripts but instead, write functions.

5. Be descriptive

descriptively_name_everything_and_do_it_early_refactor_if_necessary()

🧵

1. Use git (use GitHub)

Every analysis I've done in 2025 is a repo. GAME CHANGER.

2. Namespace everything

Every function should include pkg::fun. EVERY FUNCTION.

3. Use air formatter

Automatic formatting. IT DOES SAVE TIME.

🧵

There's a few rules I've adopted for data analysis this year that have helped me out a lot.

Friendly reminder, after 2038 integer representations of datetime are larger than 32-bits #POSIXct

Darth Vader "I've been waiting for you S7. We meet again at last" meme

When I left you. I was but the learner. Now I am the master.

Looks like David Smith & Hong Ooi 🤗 at least on the internet archive

No that's 💤.R

duckplyr is also designed like this 😍

What do you mean? Looks just like his profile pic ⚪

stRictvegetaRian

zzz.R is all you need

No it's okay. I am fine now, thanks for the offer.

As if today, I can back into the energy use of any ceiling fan in any building in America

snake_case_is_all_you_need

stringr::str_detect is all you need

I AM INVINCIBLE

YAML is all you need

left join left join left join left join

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.

🥴