#Data Availability

[kimmytsu]

squish that owl something wholesome to start the year 🥹

[risingchaos]

Daily reminder that Data glows to Geordi. That Data has an identifier no other being Geordi is in contact with has.

I wonder if it makes Data easy to spot in a crowd. If Geordi always finds himself naturally drifting towards him because he’s just.. right there. Always gold. A steady beat that doesn’t dim, that he can find with such simplicity because he stands out in the best way to him. Because everyone else is shadowed and vague and he has to carefully remember which impressionistic idea of a human is which, and Data’s ever present consistency brings him to his side time and time again. How Data is so special in so many ways, and his aura is just one more thing that makes him near perfect to Geordi.

[omg-snakes]

A sincere thanks to those of you who are sharing your snake data. Here's a little taste of what @talesfromtreatment and I have been cooking up, and this isn't even the really interesting stuff! This is only one week's worth of data capture!! Imagine what we could do with even more!

If you haven't submitted your corn snake's age, weight, and a top-down photo of them with a ruler yet, it's not too late! It's probably never going to be too late! Take the anonymous survey via Google Forms, linked below:

https://forms.gle/QtAwmiFa6fpBewFs8

Please join our body of community scientists and help us be petty in the most productive way imaginable. Your data will be used to make a gift for all corn snake keepers and breeders, present and future.

A project of this size has never before been attempted, or at least never published, and this is a super awesome opportunity for all of us to learn together!

Edit: swapped out the graph because the title said, "Width" when it should've said "Weight." Sorry!

[lyinginbedmon]

If you don't frequent my coding streams I feel like you maybe won't appreciate the last 3 hours of my life. So here's a hopefully-comprehensible explanation:

Started a new project, adding a manager AI to Minecraft villages so they can grow and expand as needed without player intervention. Call the project Hrrmowners because it's a perfect name with no issues.

Step one, naturally, is to make the infrastructure for actually understanding the layout of a village, so I make a quick system that tells me all the bounding boxes of the structures used to build villages during world generation.

Then realise that I'm actually highlighting every single structure in the game. So I need to filter things down a bit.

It takes me about an hour to find any shred of information in the disparate and convoluted world generation infrastructure of the game to have even the first clue what spawned building is part of a village. Not anything else, just if its intended to be part of one. This still catches pillager outposts, and ancient cities for some godforsaken reason.

Figure that there must be a gameplay element or two that specifically are handled by villager objects. Maybe it doesn't have all the information I need, but it might have something useful for this early stage of development.

Iron Golems spawn in villages, but it's actually the villager AI that does it, and it doesn't care about an actual village.

Cats only spawn in villages (and witch huts), and they have a spawner object in the world data. Could this be the clue I need? No, because it doesn't actually reference any kind of village data object, it just checks for occupied beds nearby via the points of interest storage of the world.

"Points of interest" are a system of noteworthy positions in the world added in 1.14, aka the Village & Pillage update, and basically let villagers reserve a bed to sleep in and a workstation to treat as its job. Every point of interest is a tile entity that gets loaded in a separate list when the chunk is restored from memory, so the world never needs to know anything about its own terrain to know where those points actually are. It just adds them in whenever the chunk spawns.

Notably removed in 1.14 is the villages.dat file in every world save. It contained the locations of every building in the village. That was fed into a Village object in world memory to monitor stuff like iron golems spawning.

It doesn't exist anymore. "Villages" as a concept, only actually exist in the world generation step, and then nothing knows or defines what they are.

TLDR: As of Minecraft 1.14, in the Village & Pillage update, the mobs named villagers actually have absolutely no clue what a village actually is.

[millenniallust4death]

Here’s the third exciting installment in my series about backing up one Tumblr post that absolutely no one asked for. The previous updates are linked here.

Previously on Tumblr API Hell

Some blogs returned 404 errors. After investigating with Allie's help, it turns out it’s not a sideblog issue — it’s a privacy setting. It pleases me that Tumblr's rickety API respects the word no.

Also, shoutout to the one line of code in my loop that always broke when someone reblogged without tags. Fixed it.

What I got working:

Tags added during reblogs of the post

Any added commentary (what the blog actually wrote)

Full post metadata so I can extract other information later (ie. outside the loop)

New questions I’m trying to answer:

While flailing around in the JSON trying to figure out which blog added which text (because obviously Tumblr’s rickety API doesn’t just tell you), I found that all the good stuff lives in a deeply nested structure called trail. It splits content into HTML chunks — but there’s no guarantee about order, and you have to reconstruct it yourself.

Here’s a stylized diagram of what trail looks like in the JSON list (which gets parsed as a data frame in R):

I started wondering:

Can I use the trail to reconstruct a version tree to see which path through the reblog chain was the most influential for the post?

This would let me ask:

Which version of the post are people reblogging?

Does added commentary increase the chance it gets reblogged again?

Are some blogs “amplifiers” — their version spreads more than others?

It’s worth thinking about these questions now — so I can make sure I’m collecting the right information from Tumblr’s rickety API before I run my R code on a 272K-note post.

Summary

Still backing up one post. Just me, 600+ lines of R code, and Tumblr’s API fighting it out at a Waffle House parking lot. The code’s nearly ready — I’m almost finished testing it on an 800-note post before trying it on my 272K-note Blaze post. Stay tuned… Zero fucks given?

If you give zero fucks about my rickety API series, you can block my data science tag, #a rare data science post, or #tumblr's rickety API. But if we're mutuals then you know how it works here - you get what you get. It's up to you to curate your online experience. XD

[chiefladylightyay]

Weiss: Penny, why did you decide on long hair for your new body?

Penny: *is quiet for a long moment* ... No data available.

Weiss: What does that even mean?

Ruby: It means she loves you. Us.

Weiss: Really? That's... Sweet. In a weird way.

Penny: *small blush* No data available.

Weiss: ... No data available. *small bashful blush*

Ruby: No data available.

[chaos0pikachu]

"why is 4mins getting a second season? who asked for this"

y'all hating b/c 4mins is that bitch, Megan Thee Stallion Yuki Chiba in Mumashi, fries dipped in a chocolate milkshake, Cher in Clueless

and like, also one of VIU's top performing shows on the platform but saying 4mins was having a hot girl summer is funnier so

[transmechanicus]

I hate online dating, the bastard dukes of Hinge have rendered me unfuckable pariah due to the Mark of Cain (openly transgender).

[theflikchic]

I never wanna see anyone say Data isn't autistic-coded ever again when he specifically became attached to Sherlock Holmes, arguably one of the most autistic-coded characters in all of fiction.

[takahikohayashi]

D-23.Jan.2018 Takahiko Hayashi 林孝彦 2019

pigment acrylic ink on paper/ original emulsion base image size: 42.2x28.0cm (Japanese Gampi paper) paper size: 54.0x39.3cm (Hahnemuhle5761)

[yeah-thats-probably-it]

thinking about what if data were aromantic. even after he gets the emotion chip put in, he still just doesn’t feel romantic love, and doesn’t understand what other people are feeling when they talk about it. and he’s really upset at first because he feels like after everything he’s been through to get here he still isn’t fully human. there’s still a piece missing. he’s tapped into every work of art humans have ever created and romance seems so ubiquitous that he thinks you can’t be human without it

i’m imagining this as like a b or c plot of an episode. maybe he’s been working with a crew member who turns out to be aromantic and they commiserate about how alienating it can be to hear everyone talk about romance all the time. everyone comes together at the end to help him realize that he feels so many different kinds of love, and no human’s experience with emotions is exactly the same and your humanity doesn’t depend on what emotions you do or don’t feel anyway

i think this would be a good episode. i would enjoy it

[folkdevilism]

Shirtless Varre brought to you by data mining.

[kangals]

oh i forgot to write up an update on the embark situation!

if you missed my posts last month (or didnt read them because they were certified walls of text): i got kep an embark test and in addition to the expected stuff, it also flagged him as having some genes related to copper storage disease - specifically that he had 1 gene for copper accumulation, and 2 genes that make accumulation less severe, meaning he inherited copper-related genes from both parents. when i did some reading into it,i couldnt find ANY studies or resources regarding copper disease in collies - it was almost entirely focused on labradors and dobermans, with an occasional hand-wavy comment of "sometimes seen in other breeds idk." because this test only got added by embark last month, i ended up doing a new test on stellina to check and see if she also had it.

basically my thinking was: A: does the presence of these genes mean that kep is actually at risk for copper storage disease, OR B: was this an incidental finding and the genes really only affect labs and dobes.

with stellina's test results, i think i'm pretty confident in guessing that the answer is B!

stellina's results also flagged as having one gene for copper accumulation - which, since i know that kep and stellina are pretty distantly related, indicates to me that it's much more likely that the genes are widespread in the greater collie gene pool as opposed to both of them just so happening to pick up a rare and harmful gene.

and most notably, now that's been a month and embark presumably has more data, they also had this new information listed:

those are HIGH numbers - and is an even bigger indicator that the gene, for whatever reason, simply doesn't affect these breeds that much. great pyrs, for example, are primarily working, livestock guardian dogs - there's no way that 80% of these dogs are suffering from copper storage disease. and while frenchies and cavs are certainly poster children for "breed with lots of serious health problems", copper storage isn't one of them. given the extreme popularity of frenchies, if 90% of them had this disease, someone would have noticed.

(interestingly, last week when i looked at this instead of pyrs it had pembroke corgis listed in that spot at around the same numbers - i wonder how many other breeds have large percentages of this gene but embark is just only listing the top few. this is such a new test that i expect a lot of information to change)

so i'm guessing that copper storage genes are actually pretty common in dogs - especially breeds with known low genetic diversity, such as collies, cavs, and dobermans. but given that there's nothing noted in studies about breeds besides labradors, dobermans, and a few terrier breeds, it just might not affect the rest of them. similarly, i know stellina carries the gene for "saddle tan" coloration, as do a certain %age of collies - but collies don't come in saddle tan. it's suspected there's another gene needed to "'activate" the pattern that's missing from collies. i'd guess the same thing is happening here - lots of dogs have the gene, but the "activation" genetics are only in a few specific breeds.

i'm definitely still going to be keeping a close eye on it, the test only released in May so it's still very early and i imagine they're getting a ton of data from this - would love some research into if my thinking is correct. if you have a dog that's gotten an embark health test from may and onwards this year, i'd be interested to know what breed(s) your dog is and if they also carry these genes. but i think, overall, this indicates that it's more incidental than a real concern.

[clonerightsagenda]

As of the end of System Collapse, SecUnit intends to travel with ART and its crew/the university rather than returning to Preservation. However, Wells is leaving some openings for Preservation characters to still appear in that context, what with Amena wanting to study at the university and Ratthi having an entanglement with one of ART's crewmembers. SecUnit is fond of Ratthi, so this may put it in the terrible position of having to actually pay attention to human love lives so it can anticipate whether Ratthi is likely to make an appearance or not.