Danielle Navarro
banner
djnavarro.net
Danielle Navarro
@djnavarro.net
not very committed to sparkle motion
the absolute worst thing about a midlife career pivot into pharmacometrics is having to live with the accursed knowledge of alcohols nonlinear PK
November 25, 2025 at 8:53 AM
omg wtaf is wrong with men???!?!! seriously lads, sort your fucking shit out i am not your mother

(yes there is a reason for this, no i am not going to elaborate, i just need to vent)
November 25, 2025 at 5:02 AM
a guy i went to high school with is suing: one of the christian brothers who ran the school groomed and sexually assaulted him. what really angers me is that everyone knew at that time. it wasn't a secret. all the kids in the school heard the story. the adults must have known, and they did nothing
November 24, 2025 at 11:01 PM
my aching middle aged back is screaming in horror
November 24, 2025 at 8:17 PM
Sigh. Yeah. I find that it's somehow easier to just shrug and accept that they do this. Pretending that men are anything other than what they are is a recipe for getting hurt. It probably makes me part of the problem, tbh. I set my expectations so low that I don't get disappointed
November 23, 2025 at 7:14 PM
ah. i'm sorry to hear it. it's a bit of a shock when you start getting those messages :/
November 23, 2025 at 6:31 PM
I used to play a little game every time I was on a long distance drive and count the number of hetero couples in cars with the woman driving vs with the man driving. The ratio was always comically lopsided.
November 23, 2025 at 1:32 AM
i mean for fucks sake twice before 9am on a rainy sunday morning is a bit much
November 22, 2025 at 10:04 PM
i really want to see the Kaplan Meier curve for the length of time a trans woman can be on the internet before a man slides into her DMs assuming she's willing to fuck him for money
November 22, 2025 at 9:56 PM
sydney is a stunningly beautiful city. it's ridiculous
November 22, 2025 at 8:34 PM
i ran to the beach this morning
November 22, 2025 at 4:17 AM
Reposted by Danielle Navarro
"The CDC is now two distinct agencies—a house divided against itself."

The staff and scientists trying to keep our basic public health infrastructure functioning are being undermined at every turn – in their work, their trust, and their safety – by these cranks, grifters, and fascists.
RFK Jr.’s CDC fills its autism webpage with anti-vaxxer talking points.

Analysis here with quotes from current and former CDC experts 👇

insidemedicine.substack.com/p/rfk-jrs-cd...
November 21, 2025 at 1:37 AM
have been experiencing a similar issue. i filed an IT request but they have not been able to rectify the problem
November 20, 2025 at 10:11 PM
This is about where I am too. I used to think TDOR was the one day of the year when cis people might actually be moved to care. I don't really believe that anymore. It's a painful day for us, but politically speaking it's meaningless. Trans lives don't matter.

xtramagazine.com/power/activi...
I’m sick of writing about dead trans people | Xtra Magazine
For 15 years, I hoped my words would change something. Now I’m grieving the world that could’ve been
xtramagazine.com
November 20, 2025 at 8:07 PM
I had noticed this recently and wasn't sure why positron does it. It hasn't caused me any problems yet but it does seem strange
November 20, 2025 at 7:59 PM
Reposted by Danielle Navarro
We absolutely do not 'need humility' to engage with AI. Technology tools need to be designed around user needs, and prioritise user outcomes. And if the tools don't do that, we should not use them. Embedding AI into everything is not a fait accompli. We are doing this to ourselves.
Concerning how folks in many fields are expressing AI fatalism—essentially arguing that like it or not, professionals in an ironclad compact with AI now and we just need to get on board
November 19, 2025 at 10:43 PM
Yeah I'll concede that's true. S3 is the most minimalist OOP system I've ever encountered, and that does make it easy to use. But, sadly, also makes it easy to mess things up when using it. Life is trade-offs I guess
November 19, 2025 at 10:42 PM
In fairness it's not my article: @clauswilke.com is the author, but I think he makes good points. The pain points he identifies with Python are the exact ones that have made me shy away from using Python as a data analysis language
November 19, 2025 at 10:40 PM
One of my favourite examples is the rxode2 package from @nlmixr2matt.bsky.social. You can specify ODE models within R cleanly, using expressions that wouldn't normally be valid R code. It make the analysts life easier, but comes at the cost of making the developers life harder
November 19, 2025 at 10:36 PM
The broader concept is "metaprogramming", and sometimes implemented in other languages as "syntactic macros". The idea is that functions can capture user code and reinterpret it. It makes it easy to write mini-languages for specific analysis tasks. Tidyverse is one example, but there are others
November 19, 2025 at 10:36 PM
It's a pity that we're stuck with S3 and S4 for so many things. Between the two of them R6 and S7 provide a really solid foundation for encapsulated OOP and functional OOP, but there's really no way to escape the older systems. S3 in particular is baked in everywhere
November 19, 2025 at 10:28 PM
today is TDOR. my mood is bleak
November 19, 2025 at 10:09 PM
Yeah. Hadley's description is excellent. NSE is the feature that lets an R function "reinterpret" the user input. In its most trivial usage it's the reason you can type library(dplyr) rather than library("dplyr"), in a more serious application it's why tidyverse works
November 18, 2025 at 9:07 PM
I know a lot of developers who truly despise R for its use of NSE, but I think this is one of those situations where R correctly pushes the pain onto the developer in order to make workflow easier for the analyst. YMMV
November 18, 2025 at 8:29 PM
As much as I'm loathe to enter into "language wars" style commentary, I think I agree with @clauswilke.com here. The lack of NSE and native missing values is really bad for Python as a data science language (as, in fairness, is the unhinged OOP situation in R)
Python is not a great language for data science. Part 2: Language features
It may be a good language for data science, but it’s not a great one.
blog.genesmindsmachines.com
November 18, 2025 at 8:25 PM