Compilation time depends on data set size ? - Is this a typo in this tutorial?


#1

Hi,

I am reading this tutorial : https://ourcodingclub.github.io/2018/04/30/stan-2.html

and I noticed the following sentence there, is the word compile there really what it should be ?

Does the last sentence in this text fragment make sense ? If yes, how ?

Stan models can take a long time to compile. One of their key advantage is that you have a lot of freedom to specify the priors (your prior knowledge and expectations of how a given parameter will behave) for your model parameters and you can really account for the complex structure your data might have. There is a time cost to this, as we’ve seen above. One way to approach this is to first make sure your code works on a small chunk of your data, and only afterwards, you can start running the code on the full data (and do other things while you wait for the models to (???) compile (???)).

Cheers,

Jozsef


See snapshot of page below :


#2

That’s not a thing, they probably meant ‘run’.


#3

Yes, thanks, it got me a bit confused :)