And water. So much water.

No one is exposed and well paid 🤔

Photo of a large rock face comprised of many thin layers. The photo shows jagged layers overlapping other layers, looking like large, pointy saw teeth. The rock face is grey with some lighter streaks.

Another area of seaside cliffs near my home town in #Newfoundland, Canada. #geology #hike #trail

Oh oh, and the gray cat trio of Chopard 1, Chopard 2, and Chopard 3. I initially thought they were a single cat until I saw the three of them sitting together on the fence 🤣

A neighbor had a cat accident a few years ago and the free range population got huge until folks helped with spay/neuter. Some unsocialized visitors include: Omega; Audemars (RIP) and her sister Piguet; the brothers Vacheron and Constantin.

All the cats are named after watches. The current indoor crew: Panerai, Seiko, Moser, and Purnell. I’m also working on socializing an outdoor cat (Oris); pretty sure he is Purnell’s dad.

Resurrecting habitat fragmentation as a process over time Habitat loss and fragmentation occur over time. Despite this truism, understanding the effects of habitat fragmentation has, in recent years, predominantly focused on interpreting snapshots of current...

Excited to share a new article from the lab & wonderful co-authors.

We argue that habitat fragmentation research needs to be recalibrated to focus on changes over time.

We hope that doing so will advance understanding, isolate attribution, and diminish heated debate.

www.cell.com/trends/ecolo...

A VERY LARGE pinkish tardigrade on green mossy surface, all from cake.

Another shot of the cake, plus info tags.

Gotta show you this tardigrade cake I saw at the American Museum of Natural History.

Oh Wizard! Sounds like a character!

I’d heard that seal point are talkative, but I had no idea 🤣

Right?!

Actually, he doesn’t! Which makes him all the more charming. And super clumsy. Diligently does multiple perimeter patrol circuits to all windows every day (and night), then reports back that all is safe.

A total babe.

The 💰 was the first thing I thought of. Well, second thing. First thing was imagining being completely covered by 10 sleeping cats <wistful sigh>

PS - This gives insight into your talent for building community among nerds

Never ever!

😵

Seal point cat sitting on the back of a chair and chewing on a bamboo stick with a charming string attached

This is Panerai. He talks. A lot.

What a goth queen

Who said anything about shame?! My feelings are awe, respect, and admiration. Ok, and probably some envy because I only have 4.

Screenshot of grid of some of my quarto extensions in a grid layout

If you are doing any #quarto slidecrafting, i have been trying really hard to keep this page updated with everything I have done
emilhvitfeldt.com/project/slid...

So if the math is mathing…you have 10 cats?!!

Fraction of estimated coefficients that are significantly different from 0 (given by the size of the colored dots), for different analysis methods (columns) and variable combinations (rows). We show three different sampling intervals (x-axis), where 1 means the smallest time interval (highest temporal resolution or shortest sampling interval) and 100 the lowest, and different degrees and combinations of autocorrelation in the landscape (y-axis). Note, we expect a positive coefficient for elevation (+2) and a negative coefficient (−2) for habitat feature

Full figure:

And this is a clever way to label circle size:

Schematic of the implemented analytical methods and simulation study of scenario 1. (A) For each of the replicated landscape settings with different spatial autocorrelation and their combinations, we (B) simulated 20 repetitions of nearly-continuous animal movement tracks influenced by these variables and subsampled the tracks to three different sampling intervals. (C) The tracks of different intervals were analyzed with three method variants (iSSA, RSA, and wRSA).

Raising the bar for Figure 1 like:

Addresses an emerging gap in wildlife monitoring + you KNOW the #dataviz is going to be pretty when @cedricscherer.com is a coauthor 💅

Absolute truth

‘I’ve been a bit shocked at how many technical terms have drifted in meaning in ML’

Not honest or clueless mistakes.
😵 e.g., anything with autoregressive structure is “causal” in a “past causes the future” sense. EVEN arbitrary directed graphical models when no explicit causality is modeled.

Super clean and the color shift is a very clear way to call out the change.

Really nice one. As usual, excellent annotation and descriptive text (i.e., saved to my inspo collection)

3/ Without a framework, it is hard to know which differences matter and why.

Formal models force you to state your assumptions explicitly. No hiding behind vague qualifications. No hand-waving. Every assumption is visible, every logical step is checkable