Matthijs Hollanders

First blog post! Stan can be daunting for ecologists because discrete parameters are not allowed with Hamiltonian Monte Carlo (HMC), and many ecological models are constructed as hidden Markov models (HMMs) where the latent ecological states of interest are discrete. (1/3)

@KrownCryptoCave Bro stop gpt generating your posts

CPI just came in at 3.3%, up from 2.4% last month. The stat that matters: gasoline surged 21.2% in March. That's the steepest single-month jump since the series began in 1967. Not 2008. Not 2020. 1967. Meanwhile BTC is holding $72K, SPX reclaimed 6800, and oil is back knocking on $99 after a 16% crash on Monday the biggest since April 2020. Core CPI only rose 0.2%. Strip out energy and inflation is largely contained. The entire CPI spike is geopolitical. That's a very different setup than "inflation is back." Pay attention to the composition, not just the headline number.

@Silimund @bratton what

@realHollanders @bratton You are in "energy crysis" becouse you decided to be in energy crysis.

"Data centers use all the water" is "vaccines cause autism" for people with graduate degrees

@digijordan @drew4worldruler One that poorly understands his craft it would seem.

@drew4worldruler I’m a photographer

Not one star. 60 years of technological progress with cameras… And not one star. I’ve heard all the explanations for this… None make any sense.

@andyroyle_pwrc Cheers!

Looks good, top of my reading list!

@bratton Sorry I don't follow. If AI uses ungodly amount of power, and we are in an energy crisis, it's a reasonable critique right?

@realHollanders that's a much more important issue, but one that is symptomatic of decades of decline in our capacity to build and renew public infrastructure, not because of compute.

...I also introduce a method to account for unequal survey intervals in the entry process, using the survey intervals as offsets in the Dirichlet/logistic-normal entry probabilities.

New pre-print! I cover a range of open capture-recapture models (single survey/robust design, multistate/multievent, in (Cormack-)Jolly-Seber variants) in Stan, and provide efficient log likelihood functions... 1/n ecoevorxiv.org/repository/vie…

@MoKetchups There was a YouTube channel called HistoryFund talking about the Big 3 and funding both sides of major conflicts. It got taken down.

This really isnt hard to see Its mapped You dont have to take the grifters word for it... you can see who paid for the words being spoken

Today's Update: 2027 and 2032 Prediction Updates In short. Full market crash by 2032 and a flip to Joulework Joulework and 0 person companies (a.k.a the Roemmelle griftconomy) become mainstream by 2027 Bookmark this

@YusakuHoriuchi Ask them for compensation. Why would you work for free?

I have already reviewed 6 manuscripts this month. I have 5 more reviews that I agreed to submit within a few weeks. I received two new requests to review yesterday. Something must be changed in the journal review system.

@rlmcelreath Nice! I’ve actually moved to student-t priors on global variance with Dirichlet or logistic normal decomp. R2D2 doesn’t make much sense for GLMs (eg, low R2 with small intercept Poisson models).

@realHollanders They will be in 3rd edition yes. Plan is to revise overfitting chapter with R2D2. I want to do a lot more priors on higher levels in other chapters too.

Statistical Rethinking 2026 is done: 20 new lectures emphasizing logical & critical statistical workflow, from basics of probability to causal inference to reliable computation to sensitivity. It's all free, made just for you. Lecture list & links: #calendar--topical-outline" target="_blank" rel="nofollow noopener">github.com/rmcelreath/sta…

@RexDouglass Yes. I remember an old post of yours saying it’s a long list of checks you need to go through and if you fuck up one of them the entire analysis is essentially worthless.

The paradox of how PhDs are some of the smartest people in the world but also their empirical work is a Pacific Garbage Patch of trash is resolved by just realizing the work is actually harder to do well than anybody is taught or pretends. It looks easy but is brain surgery.

@ryancbriggs And Gelman to you as well

@DingoResearch Same...

@realHollanders 🤣 - A very simple one: me.

Eucalyptus michaeliana- the Hillgrove Gum. Another spectacular local that is great as a street tree or for creating a shady spot in a big backyard.

@DingoResearch Very nice. I just spread out 20 cubes of wood chip for a "food forest" style orchard. Still exploring options for the regen side of the project.

@realHollanders Two different strategies in place. 1) Areas that are mulched out and planted en masse and 2) others where, as you note, individual trees are spaced out and mulched individually for a future ‘woodland’.

Another day = another Australian plant. This is Casuarina cunninghamiana Often called river oaks but grow well in drier settings. Can’t beat the sound of wind blowing through them.

@luckychartape Somehow some of this still feels like semantics. You're not predicting what price will do, but you do predict that the long term PnL of your trading process is positive.

Most traders get trapped for years, thinking that trading is about predicting what the market will do next, instead of assessing what the market is doing now. I hope this video helps many of you understand what the difference is, and how you can apply that concept to your own trading. $BTC #Bitcoin

@BigCheds @cantonmeow He’s wrong tho

@cantonmeow Interesting

#Bitcoin touching the bottom of the 2 week Keltner channel indicates that either it has already bottomed, 2% away from being bottom, or 12% from being bottomed.