Skip to content

Instantly share code, notes, and snippets.

View devmotion's full-sized avatar

David Widmann devmotion

View GitHub Profile
devmotion / Manifest.toml
Created June 25, 2023 10:57
Enzyme segfaults with Turing
View Manifest.toml
# This file is machine-generated - editing it directly is not advised
julia_version = "1.9.1"
manifest_format = "2.0"
project_hash = "65efc28519b1ed70299db030b3148e218b921e6a"
git-tree-sha1 = "dcfdf328328f2645531c4ddebf841228aef74130"
uuid = "47edcb42-4c32-4615-8424-f2b9edc5f35b"
version = "0.1.3"
devmotion / log.txt
Created September 28, 2022 10:38
Turing Enzyme
View log.txt
112,0,0,16,22]:Integer, [112,0,0,16,23]:Integer, [112,0,0,16,24]:Integer, [112,0,0,16,25]:Integer, [112,0,0,16,26]:Integer, [112,0,0,16,27]:Integer, [112,0,0,16,28]:Integer, [112,0,0,16,29]:Integer, [112,0,0,16,30]:Integer, [112,0,0,16,31]:Integer, [112,0,0,16,32]:Integer, [112,0,0,16,33]:Integer, [112,0,0,16,34]:Integer, [112,0,0,16,35]:Integer, [112,0,0,16,36]:Integer, [112,0,0,16,37]:Integer, [112,0,0,16,38]:Integer, [112,0,0,16,39]:Integer, [112,0,0,16,40]:Integer, [112,0,0,24]:Integer, [112,0,0,25]:Integer, [112,0,0,26]:Integer, [112,0,0,27]:Integer, [112,0,0,28]:Integer, [112,0,0,29]:Integer, [112,0,0,30]:Integer, [112,0,0,31]:Integer, [112,0,0,32]:Integer, [112,0,0,33]:Integer, [112,0,0,34]:Integer, [112,0,0,35]:Integer, [112,0,0,36]:Integer, [112,0,0,37]:Integer, [112,0,0,38]:Integer, [112,0,0,39]:Integer, [112,0,0,40]:Integer, [112,0,0,41]:Integer, [112,0,0,42]:Integer, [112,0,0,43]:Integer, [112,0,0,44]:Integer, [112,0,0,45]:Integer, [112,0,0,46]:Integer, [112,0,0,47]:Integer, [112,0,0,48]:Integer,
View logsumexp.birch
* One-pass algorithm of `log_sum_exp`.
function log_sum_exp_onepass(x:Real[_]) -> Real {
if length(x) > 0 {
let (mx, r) <- transform_reduce(x, (-inf, 0.0),
\(x:(Real, Real), y:(Real, Real)) -> {
let (xa, xb) <- x;
let (ya, yb) <- y;
if xa > ya {
View stacktrace.txt
┌ Warning: dt(2.220446049250313e-16) <= dtmin(2.220446049250313e-16) at t=4.2228999661198057e-16. Aborting. There is either an error in your model specification or the true solution is unstable.
└ @ SciMLBase /home/david/.julia/packages/SciMLBase/koNdH/src/integrator_interface.jl:366
Sampling 100%|██████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| Time: 0:00:09
ERROR: DimensionMismatch("arrays could not be broadcast to a common size; got a dimension with lengths 120 and 6120")
[1] _bcs1
@ ./broadcast.jl:516 [inlined]
[2] _bcs
@ ./broadcast.jl:510 [inlined]
[3] broadcast_shape
View debug.txt
julia> using DynamicPPL
julia> @model function demo4(n, ::Type{TV}=Vector{Float64}) where {TV}
m ~ Normal()
x = TV(undef, n)
@show __varinfo__
for i in eachindex(x)
x[i] ~ Normal(m, 1.0)
View discrete1d_benchmarks.jl
using BenchmarkTools
using Distances
using Distributions
using StatsBase
using LinearAlgebra
using SparseArrays
function f(
c, μ::DiscreteNonParametric, ν::DiscreteNonParametric, plan::SparseMatrixCSC
devmotion / ot.jl
Created May 24, 2021 19:09
Discrete OT
View ot.jl
using Distributions
using SparseArrays
using LinearAlgebra
using StatsBase
function _ot_cost_plan(c, μ::DiscreteNonParametric, ν::DiscreteNonParametric; get=:plan)
len_μ = length(μ.p)
len_ν = length(ν.p)
wi = μ.p[1]
wj = ν.p[1]
View sde.jl
using AdvancedMH
using ArraysOfArrays
using CairoMakie
using DiffEqNoiseProcess
using Distributions
using StochasticDiffEq
using Turing
using Random
struct CrankNicolsonProposal{P,T} <: AdvancedMH.Proposal{P}
View emcee.ipynb
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
View processes.jl
using Turing
using Turing.RandomMeasures
using Random
function stickbreaking(rpm = DirichletProcess(0.25))
# Data
data = [-2,2,-1.5,1.5]
# Base distribution