Comments on: A monad tutorial for Clojure programmers (part 4)

By: Monad Monad Monad … | 따라쟁이

Monad Monad Monad … | 따라쟁이 — Sat, 01 Sep 2012 13:53:10 +0000

[...] A monad tutorial for Clojure programmers Part 1, Part 2, Part 3, Part 4 [...]

By: khinsen

khinsen — Mon, 04 May 2009 12:43:52 +0000

Part 2 of Jim Duey’s monad tutorial also covers monad transformers, but very differently. Check it out at http://intensivesystems.net/tutorials/monads_201.html

By: khinsen

khinsen — Mon, 04 May 2009 12:43:41 +0000

You can do a lot with sequences, but I don’t see how you would do ancestral sampling. In fact, ancestral sampling is an example for what monads are made for: composing computations in which each step can depend on the result of previous steps.

As for moving around in a parameter space, I haven’t done this yet in Clojure, but as a first attempt I would put the current point in parameter space into the state of a state monad, along with the state of the random number generator.

By: Anand Patil

Anand Patil — Thu, 30 Apr 2009 14:33:00 +0000

I understand it now, thanks for the explanation! This is a very neat way of handling probabilistic operations when all of the possible outcomes can be enumerated. I can also make the leap to the ‘sampling monad’ which, I’m guessing, represents distributions by samples from them but also uses ancestral sampling for composition and a guard for conditioning.

However, one could make functionality like what you’ve shown using non-monadic tools from the sequence library (such as ‘for’ and ‘filter’). Have you ever seen the composability of monads used to embed probabilistic manipulations in a larger computation?

Also, have you given much thought to monadic expressions of algorithms that move around in the parameter space, such as maximum likelihood and MCMC?

By: A monad tutorial for Clojure programmers (part 3) « On Clojure

A monad tutorial for Clojure programmers (part 3) « On Clojure — Fri, 24 Apr 2009 16:15:48 +0000

[...] Part 4 will show you how to define your own monads by combining monad building blocks called monad transformers. As an illustration, I will explain the probability monad and how it can be used for Bayesian estimates when combined with the maybe-transformer. Possibly related posts: (automatically generated)jQuery is a MonadRemoving default values from CakePHP resultHow you should(n’t) use Monad Comments (8) [...]