StochasticAD.jl

StochasticAD.jl copied to clipboard

Differentiating a continuous time Markov chain with respect to the soujourn times

5

First issue I've done in my whole life Edit: first line of code was not visible because I don't knon Markdown This code simulates a continuous time Markov chain (CTMC)...

AndresCenteno

Stochastic triple `getindex` rule does not support vector-valued array elements

2

Hi, It appears to me that there is no support for higher dimensional random walks. Is this an inherent property of the approach or is there a method of making...

Eemilp

Performance compared to ForwardDiff

7

In this small example: ```julia using ForwardDiff, StochasticAD, BenchmarkTools function test(x, y, alpha) x = x * alpha return sum(x .* y) end x = rand(10000); y = rand(10000); alpha...

arnauqb

Add `ifelse`

2

Essentially, `ifelse(c, x, y) == [x, y][1 + !c]`

mschauer

This pull request changes the compat entry for the `Dictionaries` package from `0.3` to `0.3, 0.4`. This keeps the compat entries for earlier versions. Note: I have not tested your...

github-actions[bot]

Would be a convenience that would help avoid `StochasticAD.propagate`-based workarounds in some cases.

gaurav-arya

Add random scattering example

1

This should work as it is, but we might want to wait until more automatic solution is added, and clean up implementation before finally merging

Moelf

method error with Distributions.BinomialGeomSampler when implementing new particle filter

10

Hi! I'm trying to modify the particle filter example for a simple SIR model in discrete time (a Markov chain). When using ForwardDiff.gradient, I get the following MethodError: ``` ERROR:...

slwu89

Currently we just use the constructor, which has some issues and has become type instable for `SmoothedFIs` in some cases.

gaurav-arya

PMF of Bernoullis

1

Hey! I am trying to do VI with models involving discrete RVs. For that purpose it would be quite handy to get derivative estimators for PMFs of Bernoulli (and other...

FHoltorf