diff --git a/Project.toml b/Project.toml index ce6573a..3be77ad 100644 --- a/Project.toml +++ b/Project.toml @@ -1,6 +1,6 @@ name = "Dolo" uuid = "9d24351c-2990-5e1b-a277-04c4b809c898" -version = "0.4.2" +version = "0.4.3" [deps] AxisArrays = "39de3d68-74b9-583c-8d2d-e117c070f3a9" @@ -19,28 +19,7 @@ NLsolve = "2774e3e8-f4cf-5e23-947b-6d7e65073b56" Optim = "429524aa-4258-5aef-a3af-852621145aeb" Printf = "de0858da-6303-5e67-8744-51eddeeeb8d7" QuantEcon = "fcd29c91-0bd7-5a09-975d-7ac3f643a60c" -REPL = "3fa0cd96-eef1-5676-8a61-b3b8758bbffb" StaticArrays = "90137ffa-7385-5640-81b9-e52037218182" StringDistances = "88034a9c-02f8-509d-84a9-84ec65e18404" Test = "8dfed614-e22c-5e08-85e1-65c5234f0b40" YAML = "ddb6d928-2868-570f-bddf-ab3f9cf99eb6" - -[compat] -AxisArrays = "0.4" -BasisMatrices = "0.6.0" -Compat = "3.15.0" -DataStructures = "0.18.2" -Distributions = "0.23.8" -Dolang = "3.2.0" -HTTP = "0.8.17" -IterTools = "1.3.0" -IterativeSolvers = "0.8.4" -MacroTools = "0.5.5" -NLsolve = "4.4.1" -Optim = "0.22.0" -QuantEcon = "0.16.2" -StaticArrays = "0.12.4" -StringDistances = "0.8.0" -YAML = "0.4.2" -julia = "1.3" -Lerche = "0.4.1" \ No newline at end of file diff --git a/examples/notebooks/AR1.ipynb b/examples/notebooks/AR1.ipynb new file mode 100644 index 0000000..bb1ff35 --- /dev/null +++ b/examples/notebooks/AR1.ipynb @@ -0,0 +1,1104 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "source": [ + "# Example of an AR(1) process" + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 41, + "source": [ + "using Distributions\n", + "using Plots\n", + "using Dolo\n", + "using Test" + ], + "outputs": [], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "********** Case 1 : $X \\in \\mathbb R^2$, $A \\in \\mathcal M(2,2)$, $B \\in \\mathcal M(2,2)$ and $\\epsilon \\mathcal N(0,I_2)$ **********" + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 121, + "source": [ + "A = [0.5 1; -0.2 0.1]\n", + "B = [1 2; -1 1]" + ], + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "2×2 Matrix{Int64}:\n", + " 1 2\n", + " -1 1" + ] + }, + "metadata": {} + } + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 122, + "source": [ + "function simulate_next(Xt)\n", + " return A*Xt+B*rand(Normal(0,1),(2,1))\n", + "end\n", + "\n", + "function simulate_evolution(X0,T)\n", + " evolution = X0'\n", + " Xt=X0\n", + " for k in 1:T \n", + " Xt = simulate_next(Xt)\n", + " evolution=vcat(evolution,Xt')\n", + " end\n", + " return evolution\n", + "end" + ], + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "simulate_evolution (generic function with 2 methods)" + ] + }, + "metadata": {} + } + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 123, + "source": [ + "evolution = simulate_evolution(reshape([1.; 2.],2,1),100)\n", + "p1 = plot(evolution[:,1],evolution[:,2],xaxis = \"x1\",yaxis=\"x2\",legend = false)\n", + "p2 = plot(evolution[:,1],xaxis = \"t\",yaxis=\"x1\",legend = false)\n", + "p3 = plot(evolution[:,2],xaxis = \"t\",yaxis=\"x2\",legend = false)\n", + "plot(p1,p2,p3)" + ], + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/html": [ + "\n", + "\n", + "\n", + " \n", + " \n", + " \n", + "\n", + "\n", + "\n", + " \n", + " \n", + " \n", + "\n", + "\n", + "\n", + " \n", + " \n", + " \n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + " \n", + " \n", + " \n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + " \n", + " \n", + " \n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n" + ], + "image/png": "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", + "image/svg+xml": "\n\n\n \n \n \n\n\n\n \n \n \n\n\n\n \n \n \n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n \n \n \n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n \n \n \n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n" + }, + "metadata": {} + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "********** Case 2 : $X \\in \\mathbb R^{n*2}$, $A \\in \\mathcal M(n,n)$, $B \\in \\mathcal M(n,n)$ and $\\epsilon$ follows a $\\mathcal N(0,I)$ **********" + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 124, + "source": [ + "n=10\n", + "A = rand(n,n)\n", + "B = rand(n,n)" + ], + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "10×10 Matrix{Float64}:\n", + " 0.00539405 0.514499 0.996277 0.943604 … 0.726908 0.238462 0.335444\n", + " 0.19843 0.823938 0.326329 0.66283 0.720871 0.673999 0.116961\n", + " 0.806389 0.0967398 0.172833 0.534032 0.30306 0.933931 0.739034\n", + " 0.00731286 0.587207 0.484446 0.264356 0.732102 0.584374 0.722265\n", + " 0.792924 0.143684 0.3061 0.770247 0.272463 0.387068 0.636593\n", + " 0.246561 0.737662 0.381075 0.123529 … 0.068996 0.383876 0.937732\n", + " 0.993838 0.718556 0.965863 0.376325 0.9908 0.72653 0.466376\n", + " 0.630313 0.985313 0.900209 0.296163 0.0966201 0.934464 0.342436\n", + " 0.989114 0.117116 0.104256 0.429388 0.740371 0.114965 0.64859\n", + " 0.813805 0.60981 0.46623 0.381592 0.523726 0.381289 0.103817" + ] + }, + "metadata": {} + } + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 136, + "source": [ + "function simulate_next(Xt,n)\n", + " return A*Xt+B*rand(Normal(0,1),(n,2))\n", + "end\n", + "\n", + "function simulate_evolution(X0,T,n)\n", + " evolution = reshape(X0,1,n,2)\n", + " Xt=X0\n", + " for k in 1:T \n", + " Xt = simulate_next(Xt,n)\n", + " evolution = vcat(evolution,reshape(Xt,1,n,2))\n", + " end\n", + " return evolution\n", + "end" + ], + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "simulate_evolution (generic function with 2 methods)" + ] + }, + "metadata": {} + } + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 138, + "source": [ + "evolution2 = simulate_evolution(rand(n,2),15,n)" + ], + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "16×10×2 Array{Float64, 3}:\n", + "[:, :, 1] =\n", + " 0.708765 0.750458 … 0.862971 0.235297\n", + " 3.90591 3.8639 2.36343 1.1924\n", + " 18.9153 18.4057 14.6336 15.538\n", + " 94.0996 80.4009 75.1756 83.4702\n", + " 507.511 448.945 409.238 452.169\n", + " 2769.16 2447.16 … 2220.47 2453.24\n", + " 15048.0 13308.1 12057.5 13329.4\n", + " 81773.2 72333.9 65547.0 72451.6\n", + " 4.44468e5 3.93141e5 3.5626e5 3.9379e5\n", + " 2.41574e6 2.13678e6 1.93631e6 2.14031e6\n", + " 1.31299e7 1.16137e7 … 1.05242e7 1.16329e7\n", + " 7.13631e7 6.31225e7 5.72004e7 6.32266e7\n", + " 3.8787e8 3.4308e8 3.10893e8 3.43647e8\n", + " 2.10813e9 1.8647e9 1.68975e9 1.86777e9\n", + " 1.1458e10 1.01349e10 9.18406e9 1.01516e10\n", + " 6.22761e10 5.50848e10 … 4.99168e10 5.51757e10\n", + "\n", + "[:, :, 2] =\n", + " 0.563837 0.920074 … 0.573386 0.547473\n", + " 0.0352517 -0.143433 1.08098 1.00704\n", + " 7.19089 4.56766 5.98998 6.66668\n", + " 34.0812 29.1641 25.519 29.1668\n", + " 183.331 162.964 147.968 162.756\n", + " 1000.64 882.693 … 798.807 882.488\n", + " 5422.34 4795.94 4347.7 4805.37\n", + " 29481.0 26077.0 23634.1 26121.4\n", + " 1.60242e5 1.41735e5 1.28438e5 1.4197e5\n", + " 8.70926e5 770353.0 698078.0 7.71624e5\n", + " 4.73361e6 4.18699e6 … 3.79418e6 4.19391e6\n", + " 2.57279e7 2.2757e7 2.06219e7 2.27945e7\n", + " 1.39835e8 1.23688e8 1.12083e8 1.23892e8\n", + " 7.60026e8 6.72262e8 6.09191e8 6.73371e8\n", + " 4.13086e9 3.65385e9 3.31105e9 3.65988e9\n", + " 2.24519e10 1.98592e10 … 1.79961e10 1.9892e10" + ] + }, + "metadata": {} + } + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 146, + "source": [ + "p1 = plot(evolution2[:,:,1],xaxis = \"t\",yaxis=\"x1\",legend = false)\n", + "p2 = plot(evolution2[:,:,2],xaxis = \"t\",yaxis=\"x2\" , legend = false)\n", + "plot(p1,p2,layout = (2,1))" + ], + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/html": [ + "\n", + "\n", + "\n", + " \n", + " \n", + " \n", + "\n", + "\n", + "\n", + " \n", + " \n", + " \n", + "\n", + "\n", + "\n", + " \n", + " \n", + " \n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + " \n", + " \n", + " \n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n", + "\n" + ], + "image/png": "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", + "image/svg+xml": "\n\n\n \n \n \n\n\n\n \n \n \n\n\n\n \n \n \n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n \n \n \n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n" + }, + "metadata": {} + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "# Testing the possibility to improve simulate\n", + "\n", + "L. 306 in processes.jl, see if it is possible to get the result faster with static vectors." + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 6, + "source": [ + "using StaticArrays\n", + "using QuantEcon\n", + "using AxisArrays\n", + "using BenchmarkTools" + ], + "outputs": [], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "The next 2 chunks give examples of how to make the code evolve from matrix multiplication to static matrix multiplication." + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 218, + "source": [ + "y0 = SVector{4}(0.5,0.6,0.3,0.2)\n", + "ϵ = SVector{2}(0.5,0.4)\n", + "\n", + "R = SMatrix{4,4}(0.1, -0.2, 0.3, -0.4, 0.2, 0.1, -0.3, -0.2, 0.1, 0.2, 0.3, 0.4, -0.6, 0.3, 0.3, -0.1)\n", + "s = SMatrix{4,2}(0.1, 0.2, 0.3, 0.4, 0.1, 0.5, -0.3, 0.4)\n", + "\n", + "t0 = time()\n", + "y0 = R*y0 + s*ϵ\n", + "t1 = time()\n", + "t1-t0" + ], + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "0.0003800392150878906" + ] + }, + "metadata": {} + } + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 186, + "source": [ + "y0 = [0.5,0.6,0.3,0.2]\n", + "ϵ = [0.5,0.4]\n", + "\n", + "R = [0.1 -0.2 0.3 -0.4; 0.2 0.1 -0.3 -0.2; 0.1 0.2 0.3 0.4; -0.6 0.3 0.3 -0.1]\n", + "s = [0.1 0.2; 0.3 0.4; 0.1 0.5; -0.3 0.4]\n", + "\n", + "t0 = time()\n", + "y0 = R*y0 + s*ϵ\n", + "t1 = time()\n", + "t1-t0\n" + ], + "outputs": [ + { + "output_type": "display_data", + "data": { + "text/plain": [ + "0.00020599365234375" + ] + }, + "metadata": {} + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "Now let's describe the performance of the current function simulate and compare it with a new function called simulate_test." + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 31, + "source": [ + "@benchmark Dolo.simulate($Dolo.VAR1(0.0, ones(3,3)), $[0.37,1,-0.1]; N=20, T=100,\n", + "stochastic=true, irf=false, e0= zeros(0))" + ], + "outputs": [ + { + "data": { + "text/plain": [ + "BenchmarkTools.Trial: 6263 samples with 1 evaluation.\n", + " Range \u001b[90m(\u001b[39m\u001b[36m\u001b[1mmin\u001b[22m\u001b[39m … \u001b[35mmax\u001b[39m\u001b[90m): \u001b[39m\u001b[36m\u001b[1m472.645 μs\u001b[22m\u001b[39m … \u001b[35m 11.143 ms\u001b[39m \u001b[90m┊\u001b[39m GC \u001b[90m(\u001b[39mmin … max\u001b[90m): \u001b[39m 0.00% … 86.69%\n", + " Time \u001b[90m(\u001b[39m\u001b[34m\u001b[1mmedian\u001b[22m\u001b[39m\u001b[90m): \u001b[39m\u001b[34m\u001b[1m641.806 μs \u001b[22m\u001b[39m\u001b[90m┊\u001b[39m GC \u001b[90m(\u001b[39mmedian\u001b[90m): \u001b[39m 0.00%\n", + " Time \u001b[90m(\u001b[39m\u001b[32m\u001b[1mmean\u001b[22m\u001b[39m ± \u001b[32mσ\u001b[39m\u001b[90m): \u001b[39m\u001b[32m\u001b[1m788.538 μs\u001b[22m\u001b[39m ± \u001b[32m961.967 μs\u001b[39m \u001b[90m┊\u001b[39m GC \u001b[90m(\u001b[39mmean ± σ\u001b[90m): \u001b[39m14.94% ± 11.26%\n", + "\n", + " \u001b[39m█\u001b[34m█\u001b[39m\u001b[39m \u001b[32m \u001b[39m\u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \n", + " \u001b[39m█\u001b[34m█\u001b[39m\u001b[39m█\u001b[32m▄\u001b[39m\u001b[39m▃\u001b[39m▂\u001b[39m▂\u001b[39m▂\u001b[39m▂\u001b[39m▁\u001b[39m▂\u001b[39m▂\u001b[39m▂\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▂\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▂\u001b[39m▂\u001b[39m▂\u001b[39m▂\u001b[39m▂\u001b[39m▂\u001b[39m▂\u001b[39m \u001b[39m▂\n", + " 473 μs\u001b[90m Histogram: frequency by time\u001b[39m 7.95 ms \u001b[0m\u001b[1m<\u001b[22m\n", + "\n", + " Memory estimate\u001b[90m: \u001b[39m\u001b[33m1.30 MiB\u001b[39m, allocs estimate\u001b[90m: \u001b[39m\u001b[33m9978\u001b[39m." + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 32, + "source": [ + "function simulate_test(var::Dolo.VAR1, x0::Vector{Float64}; N::Int, T::Int,\n", + "stochastic::Bool=true, irf::Bool=false, e0::Vector{Float64}= zeros(0))\n", + "\n", + "n = size(var.mu, 1)\n", + "x0 = SVector{n}(x0)# changed from the earlier version\n", + "XN = @MArray zeros(n, N, T)\n", + "E = zeros(n, N, T)\n", + "\n", + "if stochastic\n", + " dist = QuantEcon.MVNSampler(zeros(n), var.Sigma)\n", + " for jj in 1:N\n", + " E[:, jj, :] = rand(dist, T)\n", + " end\n", + "end\n", + "\n", + "if irf\n", + " E[:, :, 1] = repeat(e0,N,1)\n", + "end\n", + "\n", + "E = SArray{Tuple{n, N, T}, Float64, 3, n*N*T}(E)\n", + "\n", + "# Initial conditions\n", + "for i in 1:N\n", + " XN[:, i, 1] = x0\n", + " for ii in 1:T-1\n", + " XN[:, i, ii+1] = var.mu+var.R*(XN[:, i, ii]-var.mu)+E[:, i, ii]\n", + " end\n", + "end\n", + "AxisArray(XN, Axis{:V}(1:n), Axis{:N}(1:N), Axis{:T}(1:T))\n", + "end" + ], + "outputs": [ + { + "data": { + "text/plain": [ + "simulate_test (generic function with 2 methods)" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 33, + "source": [ + "@benchmark simulate_test($Dolo.VAR1(0.0, ones(3,3)), $[0.37,1,-0.1]; N=20, T=100,\n", + "stochastic=true, irf=false, e0= zeros(0))" + ], + "outputs": [ + { + "data": { + "text/plain": [ + "BenchmarkTools.Trial: 6862 samples with 1 evaluation.\n", + " Range \u001b[90m(\u001b[39m\u001b[36m\u001b[1mmin\u001b[22m\u001b[39m … \u001b[35mmax\u001b[39m\u001b[90m): \u001b[39m\u001b[36m\u001b[1m428.949 μs\u001b[22m\u001b[39m … \u001b[35m 11.785 ms\u001b[39m \u001b[90m┊\u001b[39m GC \u001b[90m(\u001b[39mmin … max\u001b[90m): \u001b[39m 0.00% … 91.64%\n", + " Time \u001b[90m(\u001b[39m\u001b[34m\u001b[1mmedian\u001b[22m\u001b[39m\u001b[90m): \u001b[39m\u001b[34m\u001b[1m577.061 μs \u001b[22m\u001b[39m\u001b[90m┊\u001b[39m GC \u001b[90m(\u001b[39mmedian\u001b[90m): \u001b[39m 0.00%\n", + " Time \u001b[90m(\u001b[39m\u001b[32m\u001b[1mmean\u001b[22m\u001b[39m ± \u001b[32mσ\u001b[39m\u001b[90m): \u001b[39m\u001b[32m\u001b[1m719.460 μs\u001b[22m\u001b[39m ± \u001b[32m882.661 μs\u001b[39m \u001b[90m┊\u001b[39m GC \u001b[90m(\u001b[39mmean ± σ\u001b[90m): \u001b[39m12.75% ± 9.70%\n", + "\n", + " \u001b[39m█\u001b[34m█\u001b[39m\u001b[32m \u001b[39m\u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \n", + " \u001b[39m█\u001b[34m█\u001b[39m\u001b[32m█\u001b[39m\u001b[39m▄\u001b[39m▃\u001b[39m▂\u001b[39m▂\u001b[39m▂\u001b[39m▂\u001b[39m▂\u001b[39m▂\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▂\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▂\u001b[39m▂\u001b[39m \u001b[39m▂\n", + " 429 μs\u001b[90m Histogram: frequency by time\u001b[39m 8.04 ms \u001b[0m\u001b[1m<\u001b[22m\n", + "\n", + " Memory estimate\u001b[90m: \u001b[39m\u001b[33m974.64 KiB\u001b[39m, allocs estimate\u001b[90m: \u001b[39m\u001b[33m11985\u001b[39m." + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "For numbers of N and T quite high (example : N=20 and T=100), the function simulate_test tends to be more efficient." + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "Let's now verify if it would be possible to improve again the speed by using only Static vectors/arrays/matrix and no more MMatrix for example. In this perspective, the functions add_initial_state and change_value are built." + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 35, + "source": [ + "\"\"\"\n", + "add_initial_state(A :: SVector{L,T}, x,k::Int64) allows to add to the SVector A, in the positions T=k and\n", + "for all N, the SVector x.\n", + "\"\"\"\n", + "function add_initial_state(A :: SVector{L,T}, x,k::Int64) where {T,L}\n", + " SVector(ntuple(i->ifelse(i == k, A[i].+x, A[i]), Val{L}()))\n", + "end" + ], + "outputs": [ + { + "data": { + "text/plain": [ + "add_initial_state" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 40, + "source": [ + "\"\"\"\n", + "change_value(A :: SVector{L,T}, t,var,E) updates A for T=t+1 and for all N with the value of X_{t+1}\n", + "assessed with the help of X_t, var and E.\n", + "\"\"\"\n", + "function change_value(A :: SVector{L,T}, t,var,E) where {T,L}\n", + " SVector(ntuple(i->ifelse(i == t+1, var.mu.+var.R.*(A[t].-var.mu).+E[:, :, t], A[i]), Val{L}()))\n", + "end" + ], + "outputs": [ + { + "data": { + "text/plain": [ + "change_value" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 18, + "source": [ + "function simulate_test3(var::Dolo.VAR1, x0::Vector{Float64}; N::Int, T::Int,\n", + " stochastic::Bool=true, irf::Bool=false, e0::Vector{Float64}= zeros(0))\n", + " \n", + " n = size(var.mu, 1)\n", + " x0 = SVector{n}(x0) \n", + " XN = SVector{T,SMatrix{n,N}}([@SMatrix zeros(n,N) for j in 1:T]) \n", + " E = zeros(n, N, T)\n", + " \n", + " if stochastic\n", + " dist = QuantEcon.MVNSampler(zeros(n), var.Sigma)\n", + " for jj in 1:N\n", + " E[:, jj, :] = rand(dist, T)\n", + " end\n", + " end\n", + " \n", + " if irf\n", + " E[:, :, 1] = repeat(e0,N,1)\n", + " end\n", + " \n", + " E = SArray{Tuple{n, N, T}, Float64, 3, n*N*T}(E) \n", + " \n", + " # Initial conditions\n", + " \n", + " add_initial_state(XN,x0,1)\n", + " for t in 1:T-1\n", + " XN=change_value(XN,t,var,E)\n", + " end\n", + " #print(XN)\n", + " #AxisArray(XN, Axis{:V}(1:n), Axis{:N}(1:N), Axis{:T}(1:T))\n", + " end" + ], + "outputs": [ + { + "data": { + "text/plain": [ + "simulate_test3 (generic function with 1 method)" + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "metadata": {} + }, + { + "cell_type": "code", + "execution_count": 19, + "source": [ + "@benchmark simulate_test3($Dolo.VAR1(0.0, ones(1,1)), $[0.37]; N=20, T=100,\n", + "stochastic=true, irf=false, e0= zeros(0))" + ], + "outputs": [ + { + "data": { + "text/plain": [ + "BenchmarkTools.Trial: 111 samples with 1 evaluation.\n", + " Range \u001b[90m(\u001b[39m\u001b[36m\u001b[1mmin\u001b[22m\u001b[39m … \u001b[35mmax\u001b[39m\u001b[90m): \u001b[39m\u001b[36m\u001b[1m40.963 ms\u001b[22m\u001b[39m … \u001b[35m58.181 ms\u001b[39m \u001b[90m┊\u001b[39m GC \u001b[90m(\u001b[39mmin … max\u001b[90m): \u001b[39m0.00% … 0.00%\n", + " Time \u001b[90m(\u001b[39m\u001b[34m\u001b[1mmedian\u001b[22m\u001b[39m\u001b[90m): \u001b[39m\u001b[34m\u001b[1m44.270 ms \u001b[22m\u001b[39m\u001b[90m┊\u001b[39m GC \u001b[90m(\u001b[39mmedian\u001b[90m): \u001b[39m0.00%\n", + " Time \u001b[90m(\u001b[39m\u001b[32m\u001b[1mmean\u001b[22m\u001b[39m ± \u001b[32mσ\u001b[39m\u001b[90m): \u001b[39m\u001b[32m\u001b[1m45.390 ms\u001b[22m\u001b[39m ± \u001b[32m 3.803 ms\u001b[39m \u001b[90m┊\u001b[39m GC \u001b[90m(\u001b[39mmean ± σ\u001b[90m): \u001b[39m2.60% ± 5.35%\n", + "\n", + " \u001b[39m \u001b[39m▁\u001b[39m▆\u001b[39m \u001b[39m \u001b[39m▆\u001b[39m▃\u001b[39m█\u001b[39m▁\u001b[39m▃\u001b[39m▃\u001b[34m█\u001b[39m\u001b[39m \u001b[39m▆\u001b[39m▃\u001b[39m▃\u001b[32m \u001b[39m\u001b[39m▃\u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m▁\u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \u001b[39m \n", + " \u001b[39m▄\u001b[39m█\u001b[39m█\u001b[39m▆\u001b[39m▇\u001b[39m█\u001b[39m█\u001b[39m█\u001b[39m█\u001b[39m█\u001b[39m█\u001b[34m█\u001b[39m\u001b[39m▆\u001b[39m█\u001b[39m█\u001b[39m█\u001b[32m▇\u001b[39m\u001b[39m█\u001b[39m▁\u001b[39m▄\u001b[39m▁\u001b[39m▆\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▄\u001b[39m▆\u001b[39m▁\u001b[39m▁\u001b[39m▄\u001b[39m▁\u001b[39m▄\u001b[39m▁\u001b[39m▄\u001b[39m▄\u001b[39m▆\u001b[39m▁\u001b[39m▄\u001b[39m█\u001b[39m▆\u001b[39m▄\u001b[39m▁\u001b[39m▄\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▄\u001b[39m▁\u001b[39m▁\u001b[39m▄\u001b[39m▁\u001b[39m▁\u001b[39m▁\u001b[39m▄\u001b[39m \u001b[39m▄\n", + " 41 ms\u001b[90m Histogram: frequency by time\u001b[39m 57.6 ms \u001b[0m\u001b[1m<\u001b[22m\n", + "\n", + " Memory estimate\u001b[90m: \u001b[39m\u001b[33m12.91 MiB\u001b[39m, allocs estimate\u001b[90m: \u001b[39m\u001b[33m107277\u001b[39m." + ] + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "metadata": {} + }, + { + "cell_type": "markdown", + "source": [ + "This is less efficient." + ], + "metadata": {} + } + ], + "metadata": { + "orig_nbformat": 4, + "language_info": { + "name": "julia", + "version": "1.6.3", + "mimetype": "application/julia", + "file_extension": ".jl" + }, + "kernelspec": { + "display_name": "Julia 1.6.3", + "language": "julia", + "name": "julia-1.6" + } + }, + "nbformat": 4, + "nbformat_minor": 2 +} \ No newline at end of file