Krishna Vishal

Build failed on latest master

5

I executed `./install.sh` to build the project and got a long compilation error. Stack trace: ```c++ -- Configuring done -- Generating done -- Build files have been written to: /home/kv/AntSimulator/build...

Feature request: allow user to pass gradient tape into sample() function

2

When sampling a Turing model using ReverseDiff and memoization in a distributed setting (MCMCDistributed()), the gradient tape is compiled `n` times if there are `n` worker processes. We can avoid...

Is this expected behavior?

2

Env: julia = 1.6.1 Setfield = 0.7.1 MWE: ```julia using Setfield struct C{S,T} i::T end c = C{Float64,Int}(1) d = C{Float64, Int}(2) Case 1: @set c.i = 2 #MethodError: no...

MethodError: no method matching bijector(::MixtureModel{Multivariate, Continuous, MvNormal, Float64})

1

MWE: ```julia using Bijectors, Distributions dist = MixtureModel(MvNormal, [(ones(2), 1), (2 .* ones(2), 1)]) x = rand(dist) b = bijector(dist) ``` Error: ``` ERROR: MethodError: no method matching bijector(::MixtureModel{Multivariate, Continuous,...

`arraydist` is showing unintended behavior.

6

```julia using Turing, ReverseDiff, LoopVectorization using ReverseDiff: @grad, TrackedArray Turing.setadbackend(:reversediff) logsumexp(x::TrackedArray) = ReverseDiff.track(logsumexp, x) @grad function logsumexp(x::AbstractArray) lse = logsumexp(ReverseDiff.value(x)) return lse, Δ -> (Δ .* exp.(ReverseDiff.value(x) .- lse),) end...

MixtureModel gives lot of numerical errors while sampling

1

@mohamed82008 ```julia using Turing, ReverseDiff, Memoization using ReverseDiff: @grad, TrackedArray Turing.setadbackend(:reversediff) Turing.setrdcache(true) function logsumexp(mat) maxima = maximum(mat, dims=1) exp_mat = exp.(mat .- maxima) sum_exp_mat = sum(exp_mat, dims=1) return maxima .+...

Making Memoization.jl threadsafe.

2

Can the following library be used to make Memoization.jl threadsafe? ThreadSafe Dict: https://github.com/wherrera10/ThreadSafeDicts.jl

Code of now playing model

6

Is it possible include code for the Google's now playing model here? Even if it's not functional.

Poisson distribution support

3

The `λ` parameter of Poisson is only defined for positive reals. But here its defined for `λ >= 0`. Is this intended? ```julia dist = Poisson(0); #no problem dist =...

Lecture videos

3

Are the lecture videos available anywhere? Apologies for raising an issue. If the discussion board is available I would've asked there.