#rstats another reason to hate excel. Regional settings cause dates to do different things on different computers. Why did you just convert half my dates incorrectly excel? 😡
Remote
danwwilson 
@danwwilson@rstats.me
mastodon 4.7.3Founder of The Data Collective & Fundraising Insights
Moderator of rstats.me
Passionate about helping NFPs to get the best value out of their data. I help charities understand how data can be used to identify strategic opportunities, and how to make best use of the information available to them.
I use #RStats to make this all happen
Plant based diet | Ride 🚲 for fun.
Brisbane | Australia
524 Followers
464 Following
30 Posts
Joined October 07, 2023
{plotcolr} author:
Session ID (IYKYK):
05d6a93a44c14e11a249e4d64a0258748867a5cb7443848d4e1cd6295bdc489e4e
Replying to
@defuneste @lwpembleton @drdcarpenter @mschmidty @milesmcbain We do a list type thing in {targets}and then have a targets factory that brings each list element into its own target.
It works pretty well for our purposes and means we’re able to save our. Config as an rds object if we want to.
Replying to
@geospacedman@mastodon.social @LeafyEricScott@hachyderm.io I was going to suggest something like this too and I’d import col’s as characters and cast columns once imported.
Replying to
@maxleibman @otfrom @cwebber I always thought of UBI as the greatest potential to give people agency over crappy employers. A safety net to try something new and allow true innovation, not just VC funded privilege innovation. Oh, geez I wish Australia had it. Not for me at the moment, but for so many others with great ideas but no safety net to live while starting.
Replying to
@pglpm@c.im @JorisMeys@mstdn.social
@Posit@fosstodon.org the R community would expect a bit better. Do better!!!
Replying to
@mdsumner@rstats.me Thanks for understanding. I appreciate you being part of the community. It’s great having you here.
Replying to
@mdsumner@rstats.me Not jail, but you never know who’s seeing your posts, and in the same way we filter our words in public, it good to do on here. I don’t moderate too much, so let’s treat it in that spirit.
@mdsumner@rstats.me I appreciate the frustration, but let’s pop a content warning on this one for those who might not like that language in their feed.
Replying to
@robyn@yeg.bike @zkamvar@hachyderm.io We do data analytics for the charity sector in Australia and don’t use gen AI. We build ML models where it makes sense, but otherwise we use human intelligence, understanding and experience to help our clients do more with their data furthering their causes. So not a unicorn, but we’re profitable enough not to need to buy into the hype and our clients trust us enough not to ask for that crap.
Replying to
@eliocamp Sorry have no cents to give. Have stuck with vscode for its SSH connectivity. Haven’t even bothered with positron yet.
Replying to
@seachanger@alaskan.social @IrrationalMethod@social.coop I think it’s a great idea and if you’re permitted to share the final policy would love to see it. I work in the NFP space in Australia and know orgs here need help with policies on this front.
Replying to
@aud@fire.asta.lgbt @seachanger@alaskan.social I was going to suggest a rewording as courts have deemed it isn’t theft to use it (also disagree with the courts) but maybe indicating that materials for training are often collected in a way that is illegal and has actively seen citizens prosecuted in the past.
So {targets} is a great package, but sometimes it creates friction when working interactively. So today, I made a little helper function to load a selection of targets. It allows me to bind this to a shortcut key in VSCode that will load all the highlighted targets.
So far, it is going very well for when I am debugging a function and need to load a bunch of targets into memory. #rStats #workflow
Replying to
@nemo@mas.to @itsfoss@mastodon.social Yeah, saw some stuff on Fedi about a new fork being taken at the last pre AI commit.
Replying to
@nemo@mas.to @itsfoss@mastodon.social Seems that may have a limited shelf life now that the developers are using AI to build it.
Replying to
@lwpembleton I have this in a package:
tar_load_selection <- function(..., store = targets::tar_config_get("store")) {
target_names <- as.character(as.list(substitute(...())))
targets::tar_load(target_names, store = store, envir = globalenv())
}
keybinding like:
{
"key": "l",
“name": "load target at cursor",
"type": "command",
"command": "r.runCommandWithSelectionOrWord",
"args": “tdctargets::tar_load_selection($$)"
},
Replying to
@gavin@fosstodon.org @jimgar@hachyderm.io @brodriguesco@fosstodon.org @rabaath@fosstodon.org @matthewbadger@fosstodon.org So it turns out I stumbled on a bug in the latest {pointblank} that was not allowing me to make things work in {targets}. @yjunechoe@fosstodon.org came to the rescue with a fix very quickly (many thanks 🙏) so I’ve put together a very basic Gist to show how it can work. https://gist.github.com/danwwilson/6d9c87187baf427950e0c07ceea50bd1 I make use of the YAML file to create a persistent ‘config’ type object.
Replying to
Replying to
@gavin@fosstodon.org @jimgar@hachyderm.io @brodriguesco@fosstodon.org @rabaath@fosstodon.org @matthewbadger@fosstodon.org Are you doing this in a script or something like {targets} :targets:?
Replying to
@yjunechoe@fosstodon.org @gavin@fosstodon.org @jimgar@hachyderm.io @brodriguesco@fosstodon.org @rabaath@fosstodon.org @matthewbadger@fosstodon.org If I hadn’t raised the issue I would have thought it was all me, so I’m just appreciative that you were able to make sense of it and fix it so quickly. 🎉
Replying to
Replying to
@isol@mastodon.au @catsalad@infosec.exchange Age verification on every container you spin up? Instead of containers for cash (can recycling) we’ll have someone running containers for kids. How to to have a privacy respecting computer
Replying to