/sci/ - Science & Math

>>12692349
>just plopping some inflatables underground and covering it with on-site martian concrete them isn't a sufficient solution

>>9520978

This entire thread is fucking disgusting. I reported each and every post.

>tfw my brother who majored in Art is 200k starting

Hello,

I'm trying to fit a mixture of factor analyzers to some high dimensional data using the R package HDclassif. hddc() uses automatic model selection according to the BIC of every possible model. Then it stores the model with the highest BIC. I can compute the AIC for that model with the likelihood, which is also stored.

But what if I want to compare different models? I only have the BIC for them, not the likelihoods, so I can't compute AIC etc.. Has anyone had this problem and knows a solution that does not involve fitting every single model one at a time?

>>4716040
>concrete cooling towers
what no why would they do that
they already know what it is, why are you wasting cg on that

everything about this movie pisses me the fuck off.
what is this about radiation poisoning causing paranoia. what the fuck. how the fuck did they get radiation poisoning strong enough to cause their skin to slough off in pripyat it isn't that fucking bad

no
no
no
no
no
THIS IS WHAT'S WRONG WITH AMERICA