Imputing new worlds with multiversal appeal

Monthly Archives: May 2016

Models many we fit,

But how to choose which it is,

We look at likelihood and the like,

One with a certain value, our interests spike.


Take the AIC where you want the lowest,

So a few ones, you put to this test.

And what may the final model help us to predict?

From which dimension should Anton Randy evict.


And because summer is coming …



So you have groups, more than two,

If they differ, what are you to do?

Maybe apply the F-statistic if need be,

Do different dimensions really differ you see?


Take Randy’s cast and what he does best.

He might have to know how to use the F-test.

To catch Anton it is just a must.

And this would be a good script at table reads, I trust.


Yeah, still thinking about doing table readings. What can you do?


Generate data you must,

But the joint distribution can’t always trust,

Still individual variables are there,

And pairwise correlations if you care.


So combine the two and see what you get,

If the algorithm converges, you’re all set.

Based on the criteria of correlations.

And if it does, we celebrate!


See? Celebrate!