Tor Erlend Fjelde torfjelde

PhD student @ Computational and Biological Learning (CBL), University of Cambridge Mathematics Bsc. student at University of Edinburgh

torfjelde / example.jl

Last active December 20, 2021 12:22

This requires `NestedSamplers#tor/improvements` and `[email protected]` or higher.

	julia> include("nested_samplers.jl")

	julia> @model function demo()
	m ~ Normal()
	x ~ Normal(m, 1)
	y ~ Normal(x, 1)
	return (; m, x, y)
	end
	demo (generic function with 2 methods)

torfjelde / turing-ad-benchmark.jl

Last active November 18, 2023 22:26

Convenient code for benchmarking different AD-backends on a particular Turing.jl model.

	# Use packages to ensure that we trigger Requires.jl.
	using Zygote: Zygote
	using ReverseDiff: ReverseDiff
	using ForwardDiff: ForwardDiff
	using Tracker: Tracker
	using Memoization: Memoization # used for ReverseDiff.jl cache.

	using Turing.Core: ForwardDiffAD, ReverseDiffAD, TrackerAD, ZygoteAD, CHUNKSIZE

	const DEFAULT_ADBACKENDS = [

torfjelde / dispatch-on-model.jl

Last active June 11, 2021 09:59

	julia> using DynamicPPL

	julia> @model function demo1(x)
	x ~ Normal()
	end
	demo1 (generic function with 1 method)

	julia> @model function demo2(x)
	x ~ Normal()
	end

torfjelde / dppl-trace.jl

Last active April 26, 2021 12:45

A simple attempt to work generate logjoint for a simple DPPL model. Requires the following branch to work: https://github.com/TuringLang/DynamicPPL.jl/tree/tor/symbolics.

	julia> using DynamicPPL, Distributions, Bijectors

	julia> using Symbolics

	julia> using Symbolics: SymbolicUtils

	julia> import StatsFuns

	julia> # Make stuff from Distributions.jl primitives so that we don't recurse into them.
	@register Normal()

torfjelde / statsfuns-overload.jl

Last active April 6, 2021 18:50

Example of fixing `poiscdf` from Statsfuns.jl

	julia> using Distributions, ForwardDiff

	julia> import Distributions: StatsFuns # StatsFuns.jl is used within Distributions.jl, so we just grab from there

	julia> ForwardDiff.derivative(λ -> StatsFuns.poiscdf(λ, 1), 1.0)
	ERROR: MethodError: no method matching poiscdf(::ForwardDiff.Dual{ForwardDiff.Tag{var"#1#2", Float64}, Float64, 1}, ::Int64)
	Closest candidates are:
	poiscdf(::Union{Float64, Int64}, ::Union{Float64, Int64}) at /home/tor/.julia/packages/StatsFuns/zJ1EI/src/rmath.jl:77
	Stacktrace:
	[1] (::var"#1#2")(λ::ForwardDiff.Dual{ForwardDiff.Tag{var"#1#2", Float64}, Float64, 1})

torfjelde / turing-new-interface.jl

Last active November 18, 2023 22:26

Simple example of using NUTS with the new iterator interface in AbstractMCMC.jl available using Turing.jl > 0.15.

	julia> using Turing, Random

	julia> @model function gdemo(xs)
	# Assumptions
	σ² ~ InverseGamma(2, 3)
	μ ~ Normal(0, √σ²)
	# Observations
	for i = 1:length(xs)
	xs[i] ~ Normal(μ, √σ²)
	end

torfjelde / tennis_data.mat

Last active November 22, 2020 16:31

Implementation of Gibbs sampling for TrueSkill in Turing.jl.

View raw

torfjelde / generated_quantities_chains.jl

Last active March 28, 2024 20:03

Converting output from `generated_quantities(model, chain)` into a `MCMCChains.Chains` object

torfjelde / advi-simple-example.jl

Created October 6, 2020 14:50

Simple example of using AdvancedVI.jl + Bijectors.jl to construct a approximate/variational posterior.

	julia> using Bijectors

	julia> using ForwardDiff

	julia> using ComponentArrays, UnPack

	julia> using Bijectors: LeakyReLU, Shift, Scale

	julia> # Affine transformation is simply a `Scale` composed with `Shift`
	using Bijectors: Shift, Scale

torfjelde / loo.jl

Last active November 18, 2023 22:26

Using`loo` in Turing.jl