Wolfram Mathematica

« first day (3408 days earlier) ← previous day next day → last day (1059 days later) »

3:17 AM

So...is 12.3 actually out yet or...? Seems like it’s a Wolfram Cloud only release so far, hey?

3:32 AM

@CATrevillian It hasn't been announced, I haven't received an email about it and the release version is not downloadable for me on the user portal. WRI has those odd rolling releases, but I'd say that usually when new versions are "released", at least one of those things has happened.

At the same time wording of some people talking about it sound like they actually have the release version, instead of just becoming more open about testing versions as WRI becomes more open about changes etc...

Definitely not the first time those releases seem to be in a strange limbo, usually for roughly a week.

4 hours later…

kirma

7:19 AM

Don't try to massage your data too much...

I tried to retrieve "convolution kernel" of effects of consuming alcohol on blood pressure and heart rate from my data. I have thousands of measurements on those vitals and I have timestamps of all alcohol I've consumed for years. Sounds good, right?

The initial result was this, and I was quite happy, although having some concerns of biases on the dataset:

But what about that nagging feeling in the back of my head? I replaced my real heart rate and blood pressure data with simple diurnal sinusoidal functions with different shifts and ran the fit on that, using original timestamps:

Makes one much less certain of what is real and what is just a side effect of measurements or activities having uneven temporal distribution...

1 hour later…

Carl Lange

8:50 AM

@kirma That nagging feeling in the back of your head might be a hangover too, though :D

Woke up this morning to an email that 90% of my cloud credits had been used. A bit rough since I was sitting on about 200k of them. Turns out one of the random little web forms I made (this one to view your local election constituency here) got super popular and ate them all

kirma

@CarlLange It's plausible... ;)

Carl Lange

@kirma I'm really jealous of your data, it sounds very fun to work with. I don't have any of the stats knowledge to do anything sensible with it though, I would definitely not have thought to replace the real data with fake and see if it comes up the same

Sometimes I think I'm happier with unkown unknowns. I'll just sit happily in my dunning-kruger bubble and make assumptions about things.

kirma

No need to be too jealous, I would say that 95% of what comes out from it is either blindingly obvious or needs to be subject of significant suspicion... but maybe it's better to have it than not. At least that's my usual take, that more data is better...

Carl Lange

@kirma Very well, jealously levels adjusted :) Still I think it's a lot of fun to have that scale of data to play with. The creativity aspect of that kind of thing is very enjoyable.

kirma

Yep!

7 hours later…