{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Lab 14: MCMC for Feynman Path Integral Tutorial" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "In the lecture we saw some results from S.Mittal et al, and we will try to reproduce them in the lab today. The Markov chain Monte Carlo (MCMC) method can be used to evaluate the\n", "discrete imaginary-time path integral (an alternative formulation to the path integral) of a quantum harmonic oscillator system with \"Euclidean Lagragian\" $L_E(x(\\tau))=\\frac{m}{2}(\\frac{dx}{d\\tau})^2 +V(x(\\tau))$, and we can obtain a histogram of the probability density by summing up the number of times a particle is in a bin. " ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import numpy as np\n", "import matplotlib.pyplot as plt\n", "import tqdm" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Analytic solution of the Schodinger equation to the ground state wavefunction of a quantum harmonic oscillator is given by: $\\psi(x)=(\\frac{m\\omega}{\\pi\\hbar})^{1/4} \\exp{(-\\frac{m\\omega}{2\\hbar} x^2)}$.\n", "\n", "A different approach to solving the Schodinger equation is by evaluating the imaginary-time path integral for this system using the Markov chain Monte Carlo (MCMC) method.\n", "\n", "\n", "The path is assigned along a time lattice to $N_{xbins} =100$ spatial bins ranging in $[-4,4]$ with width $\\Delta x$. Choose time lattice with $N_{\\tau} = 30$ time increments $\\delta\\tau$ with lattice points $x_n = n\\delta \\tau$ for $n = 0,1,2,...,N_{\\tau}$, over $n_{sweeps} = 20,000$, hitsize $h = 0.1$, and set $m=\\hbar=1=\\omega$. Periodic boundary condition is imposed on the lattice. \n" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "SWEEPS = 20_000\n", "H = 1\n", "OMEGA = 1\n", "M = 1\n", "TAU = 30\n", "DELTATAU = 1\n", "NTAU = int(TAU/DELTATAU)\n", "HITSIZE = 0.1\n", "\n", "XLOW = -4\n", "XHIGH = 4\n", "NXBINS = 100\n", "DELTAX = (XHIGH - XLOW) / NXBINS\n", "HITSIZE = 0.1\n", "prob_histogram = np.zeros(NXBINS)\n", "x_bins = np.linspace(XLOW, XHIGH, NXBINS + 1)\n", "\n", "x_path = np.zeros(NTAU)\n", "## analytical solution\n", "psi_analytical = (M*OMEGA/(np.pi*H))**0.25*np.exp(-M*OMEGA*x_bins**2/2/H)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We begin with an initial path $x_{path}$, which may be an array of random numbers (‘hot’ start) or zeros (‘cold’ start), and we use the ‘cold’ start for now. This path is updated by applying the Metropolis–Hastings algorithm to each element $x_i$ of the path in random order, called a ‘sweep’. There are two steps in the updating process:\n", "- Generate a uniform random number $u \\in [-h, h]$ (where h is called the hitsize)\n", "- Propose the new value $x_i'=x_i +u$ of the path element and calculate the resulting change $\\Delta S$ in the action.\n", "If $\\Delta S \\leq 0$, accept the new path element.\n", "If $\\Delta S > 0$, accept with probability $exp(-\\Delta S/\\hbar)$.\n", "\n", "For better performance, we can play with the parameters DELTATAU, HITSIZE, SWEEPS, etc, and we can also add \"thinning\" where only every Nth sweep is saved. For example:\n", "- if we choose a smaller time step $\\delta\\tau = 0.1$, we end up with better energy approximation to the ground state; \n", "- if the hitsize we choose is too large, few changes will be accepted; too small and the exploration of phase space will be slow. How should we choose the hit size wisely?\n", "\n", "The action is $S(\\{x_k\\})=\\delta\\tau\\sum_{i=0}^{N_{\\tau}-1}[\\frac{m}{2}(\\frac{x_{i+1}-x_i}{\\delta \\tau})^2 +V(x_i)]$, with $x_{N_\\tau} \\equiv x_0$ and $V(x) = \\frac{m\\omega^2}{2}x^2$.\n", "This evaluates $V$ at the left end of each time step, as in Mittal et al. and Assignment 3. Lecture 13 evaluates it at the midpoint $(x_i+x_{i+1})/2$ instead; both have the same $\\delta\\tau \\to 0$ limit, but for the harmonic oscillator the midpoint form happens to give the exact ground-state $\\langle x^2 \\rangle$ at any $\\delta\\tau$, which would hide the discretization error we want to study.\n", "Note that one sweep produces the next path. Each path is determined only by\n", "the immediately preceding path, so the complete sequence of paths forms a Markov chain,\n", "but the paths are correlated." ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [], "source": [ "def vary_path(x_current):\n", " x_prime = x_current + np.random.random() * 2 * HITSIZE - HITSIZE\n", " while x_prime > XHIGH or x_prime < XLOW:\n", " x_prime = x_current + np.random.random() * 2 * HITSIZE - HITSIZE\n", " return x_prime\n", "\n", "def action(x_left, x_right):\n", " K = 0.5 * M * (((x_right - x_left))**2) / DELTATAU\n", " V = 0.5 * M * DELTATAU * (OMEGA**2) * x_left**2\n", " return K + V\n", "\n", "def total_action(x_path):\n", " path_action = 0\n", " for i in range(-1, NTAU-1):\n", " path_action += action(x_path[i], x_path[i+1])\n", " return path_action\n", "\n", "def delta_action(x_path, x_prime, i):\n", " x_left = x_path[i-1]\n", " x_right = x_path[i+1] if i < NTAU-1 else x_path[0] #PBC.\n", " daction = action(x_left, x_prime) + action(x_prime, x_right) \n", " daction -= action(x_left, x_path[i]) + action(x_path[i], x_right) #compute the resulting change from u in delta S.\n", " return daction\n", "\n", "def MCMC(x_path, prob_histogram):\n", " for i in range(NTAU):\n", " x_prime = vary_path(x_path[i])\n", " daction = delta_action(x_path, x_prime, i)\n", " if daction <= 0: \n", " x_path[i] = x_prime\n", " else: \n", " prob = np.exp(-daction)\n", " if np.random.random() < prob:\n", " x_path[i] = x_prime\n", " hist, _ = np.histogram(x_path, bins=x_bins)\n", " prob_histogram += hist" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "100%|██████████| 20000/20000 [00:02<00:00, 8846.43it/s] \n" ] } ], "source": [ "for k in tqdm.tqdm(range(SWEEPS)):\n", " MCMC(x_path, prob_histogram)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Finally we get the histogram of the probability density by summing up the number of times a particle is in a bin, and compare it to the analytical solution. The MCMC histogram is close to the analytical curve but slightly narrower: with $m=\\hbar=\\omega=1$ and a long path, the discretized action gives $\\langle x^2 \\rangle = 1/\\sqrt{4+\\delta\\tau^2}$, which is $0.447$ at $\\delta\\tau = 1$ instead of $1/2$. Try $\\delta\\tau = 0.1$ to see it converge, as in Fig. 5 of Mittal et al.\n", "\n", "We can now play with the code by plotting the $\\langle E\\rangle$ observable for each sweep (or equivalently $\\langle x^2 \\rangle$ vs. iteration) to measure the equilibrium time; or measure the autocorrelation function/autocorrelation time, etc. " ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "data": { "image/png": "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", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plt.figure()\n", "plt.stairs(prob_histogram/np.sum(prob_histogram*DELTAX), x_bins, label=f\"MCMC, $\\\\tau = {TAU}$, $\\\\delta\\\\tau={DELTATAU}$\")\n", "plt.plot(x_bins, psi_analytical*psi_analytical.conjugate(), label=\"Analytical\")\n", "plt.legend()\n", "plt.show()" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.9.16" } }, "nbformat": 4, "nbformat_minor": 2 }