diff --git a/docs/source/mcmc_samplers/dram_ac_mcmc.rst b/docs/source/mcmc_samplers/dram_ac_mcmc.rst new file mode 100644 index 000000000..d157de698 --- /dev/null +++ b/docs/source/mcmc_samplers/dram_ac_mcmc.rst @@ -0,0 +1,7 @@ +********* +Dram ACMC +********* + +.. currentmodule:: pints + +.. autoclass:: DramACMC diff --git a/docs/source/mcmc_samplers/index.rst b/docs/source/mcmc_samplers/index.rst index ac75b078d..14a57796a 100644 --- a/docs/source/mcmc_samplers/index.rst +++ b/docs/source/mcmc_samplers/index.rst @@ -15,6 +15,7 @@ interface, that can be used to sample from an unknown base_classes adaptive_covariance_mc differential_evolution_mcmc + dram_ac_mcmc dream_mcmc emcee_hammer_mcmc haario_ac_mcmc diff --git a/examples/README.md b/examples/README.md index 252190933..237055940 100644 --- a/examples/README.md +++ b/examples/README.md @@ -38,6 +38,7 @@ relevant code. ### MCMC without gradients - [Differential Evolution MCMC](./sampling/differential-evolution-mcmc.ipynb) +- [DRAM ACMC](./sampling/adaptive-covariance-dram.ipynb) - [DREAM MCMC](./sampling/dream-mcmc.ipynb) - [Emcee Hammer](./sampling/emcee-hammer.ipynb) - [Haario Adaptive Covariance MCMC](./sampling/adaptive-covariance-haario.ipynb) diff --git a/examples/sampling/adaptive-covariance-dram.ipynb b/examples/sampling/adaptive-covariance-dram.ipynb new file mode 100644 index 000000000..a52efc161 --- /dev/null +++ b/examples/sampling/adaptive-covariance-dram.ipynb @@ -0,0 +1,299 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "# Inference: Delayed Rejection Adaptive Covariance MCMC\n", + "\n", + "This example shows you how to perform Bayesian inference on a time series, using [DRAM ACMC](http://pints.readthedocs.io/en/latest/mcmc_samplers/dram_ac_mcmc.html) as described in [1]. This method allows users to have a number of proposal kernels. Typically, the first proposal kernel is wider (and hence can explore more aggressively); if this proposal is rejected, then subsequent kernels are more narrower to encourage, at least, some movement of the chains.\n", + "\n", + "\n", + "[1] \"DRAM: Efficient adaptive MCMC\".\n", + " H Haario, M Laine, A Mira, E Saksman, (2006) Statistical Computing\n", + " https://doi.org/10.1007/s11222-006-9438-0" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "In our first use of the model, we try using 3 different proposal kernels that scale the proposal distributions to be of three different orders of magnitude." + ] + }, + { + "cell_type": "code", + "execution_count": 1, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Running...\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/Volumes/Samsung1.5TB/Github/pints/pints/_mcmc/_dram_ac.py:113: RuntimeWarning: invalid value encountered in double_scalars\n", + " alpha_log = log_Y[n + 1] - log_Y[0]\n", + "/Volumes/Samsung1.5TB/Github/pints/pints/_mcmc/_dram_ac.py:131: RuntimeWarning: divide by zero encountered in log\n", + " i, Y_rev[0:(i + 2)], log_Y_rev[0:(i + 2)]))) -\n", + "/Volumes/Samsung1.5TB/Github/pints/pints/_mcmc/_dram_ac.py:133: RuntimeWarning: divide by zero encountered in log\n", + " i, Y[0:(i + 2)], log_Y[0:(i + 2)])))\n", + "/Volumes/Samsung1.5TB/Github/pints/pints/_mcmc/_dram_ac.py:133: RuntimeWarning: invalid value encountered in double_scalars\n", + " i, Y[0:(i + 2)], log_Y[0:(i + 2)])))\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Done!\n" + ] + }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "import pints\n", + "import pints.toy as toy\n", + "import pints.plot\n", + "import numpy as np\n", + "import matplotlib.pyplot as plt\n", + "import time\n", + "\n", + "# Load a forward model\n", + "model = toy.LogisticModel()\n", + "\n", + "# Create some toy data\n", + "real_parameters = [0.015, 500]\n", + "times = np.linspace(0, 1000, 1000)\n", + "org_values = model.simulate(real_parameters, times)\n", + "\n", + "# Add noise\n", + "noise = 10\n", + "values = org_values + np.random.normal(0, noise, org_values.shape)\n", + "real_parameters = np.array(real_parameters + [noise])\n", + "\n", + "# Get properties of the noise sample\n", + "noise_sample_mean = np.mean(values - org_values)\n", + "noise_sample_std = np.std(values - org_values)\n", + "\n", + "# Create an object with links to the model and time series\n", + "problem = pints.SingleOutputProblem(model, times, values)\n", + "\n", + "# Create a log-likelihood function (adds an extra parameter!)\n", + "log_likelihood = pints.GaussianLogLikelihood(problem)\n", + "\n", + "# Create a uniform prior over both the parameters and the new noise variable\n", + "log_prior = pints.UniformLogPrior(\n", + " [0.01, 400, noise*0.1],\n", + " [0.02, 600, noise*10]\n", + ")\n", + "\n", + "# Create a posterior log-likelihood (log(likelihood * prior))\n", + "log_posterior = pints.LogPosterior(log_likelihood, log_prior)\n", + "\n", + "# Choose starting points for 4 mcmc chains\n", + "xs = log_prior.sample(4)\n", + "\n", + "# Create mcmc routine\n", + "mcmc = pints.MCMCController(log_posterior, len(xs), xs, method=pints.DramACMC)\n", + "\n", + "# Add stopping criterion\n", + "mcmc.set_max_iterations(2000)\n", + "\n", + "# Start adapting after 1000 iterations\n", + "mcmc.set_initial_phase_iterations(1000)\n", + "\n", + "# Disable logging mode\n", + "mcmc.set_log_to_screen(False)\n", + "\n", + "# Try 3 proposal kernels\n", + "for i in range(len(xs)):\n", + " mcmc.samplers()[i].set_n_kernels(3)\n", + "\n", + "start = time.time()\n", + "# Run!\n", + "print('Running...')\n", + "chains = mcmc.run()\n", + "print('Done!')\n", + "end = time.time()\n", + "\n", + "# Show traces and histograms\n", + "pints.plot.trace(chains)\n", + "\n", + "# Show graphs\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "We now look at the MCMC sample quantiles and sampling diagnostics." + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "param mean std. 2.5% 25% 50% 75% 97.5% rhat ess ess per sec.\n", + "------- ------ ------ ------ ------ ------ ------ ------- ------ ----- --------------\n", + "r 0.02 0.00 0.01 0.01 0.02 0.02 0.02 1.16 37.39 1.49\n", + "k 487.80 26.79 423.93 483.40 498.89 500.63 520.94 1.66 44.85 1.78\n", + "sigma 28.29 20.62 9.33 10.36 14.03 48.83 67.18 2.08 12.30 0.49\n" + ] + } + ], + "source": [ + "results = pints.MCMCSummary(chains=chains, time=end-start, parameter_names=[\"r\", \"k\", \"sigma\"])\n", + "print(results)" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "With three kernels, things look a little unstable in the sampling, so let's try just two instead." + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Running...\n" + ] + }, + { + "name": "stderr", + "output_type": "stream", + "text": [ + "/Volumes/Samsung1.5TB/Github/pints/pints/_mcmc/_dram_ac.py:131: RuntimeWarning: divide by zero encountered in log\n", + " i, Y_rev[0:(i + 2)], log_Y_rev[0:(i + 2)]))) -\n" + ] + }, + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Done!\n" + ] + }, + { + "data": { + "image/png": "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\n", + "text/plain": [ + "
" + ] + }, + "metadata": { + "needs_background": "light" + }, + "output_type": "display_data" + } + ], + "source": [ + "# Create mcmc routine\n", + "mcmc = pints.MCMCController(log_posterior, len(xs), xs, method=pints.DramACMC)\n", + "\n", + "# Add stopping criterion\n", + "mcmc.set_max_iterations(2000)\n", + "\n", + "# Start adapting after 1000 iterations\n", + "mcmc.set_initial_phase_iterations(1000)\n", + "\n", + "# Disable logging mode\n", + "mcmc.set_log_to_screen(False)\n", + "\n", + "# Try 3 proposal kernels\n", + "for i in range(len(xs)):\n", + " mcmc.samplers()[i].set_n_kernels(2)\n", + "\n", + "# Run!\n", + "print('Running...')\n", + "chains = mcmc.run()\n", + "print('Done!')\n", + "\n", + "# Show traces and histograms\n", + "pints.plot.trace(chains)\n", + "\n", + "# Show graphs\n", + "plt.show()" + ] + }, + { + "cell_type": "markdown", + "metadata": {}, + "source": [ + "The above looks much better and we typically get a higher ESS than for the three kernel case." + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": {}, + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "param mean std. 2.5% 25% 50% 75% 97.5% rhat ess ess per sec.\n", + "------- ------ ------ ------ ------ ------ ------ ------- ------ ----- --------------\n", + "r 0.02 0.00 0.01 0.01 0.02 0.02 0.02 1.22 44.99 1.79\n", + "k 497.05 26.98 429.13 495.14 499.39 501.20 539.60 2.13 43.37 1.73\n", + "sigma 25.04 19.02 9.35 10.26 11.50 42.06 69.10 1.69 13.67 0.54\n" + ] + } + ], + "source": [ + "results = pints.MCMCSummary(chains=chains, time=end-start, parameter_names=[\"r\", \"k\", \"sigma\"])\n", + "print(results)" + ] + } + ], + "metadata": { + "kernelspec": { + "display_name": "Python 3", + "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": "3.7.7" + } + }, + "nbformat": 4, + "nbformat_minor": 2 +} diff --git a/pints/__init__.py b/pints/__init__.py index 20fc710d9..472926393 100644 --- a/pints/__init__.py +++ b/pints/__init__.py @@ -195,8 +195,12 @@ def version(formatted=False): MultiChainMCMC, SingleChainMCMC, ) +# base classes first from ._mcmc._adaptive_covariance import AdaptiveCovarianceMC + +# methods from ._mcmc._differential_evolution import DifferentialEvolutionMCMC +from ._mcmc._dram_ac import DramACMC from ._mcmc._dream import DreamMCMC from ._mcmc._emcee_hammer import EmceeHammerMCMC from ._mcmc._haario_ac import HaarioACMC diff --git a/pints/_mcmc/_dram_ac.py b/pints/_mcmc/_dram_ac.py new file mode 100644 index 000000000..9a003983a --- /dev/null +++ b/pints/_mcmc/_dram_ac.py @@ -0,0 +1,290 @@ +# +# DRAM AC MC method +# +# This file is part of PINTS (https://github.com/pints-team/pints/) which is +# released under the BSD 3-clause license. See accompanying LICENSE.md for +# copyright notice and full license details. +# +from __future__ import absolute_import, division +from __future__ import print_function, unicode_literals +import pints +import numpy as np +import scipy.stats as stats + + +class DramACMC(pints.AdaptiveCovarianceMC): + """ + DRAM (Delayed Rejection Adaptive Covariance) MCMC, as described in [1]_. + + In this method, rejections do not necessarily lead an iteration to end. + Instead, if a rejection occurs, another point is proposed although + typically from a narrower (i.e. more conservative) proposal kernel than was + used for the first proposal. + + In this approach, in each iteration, the following steps return the next + state of the Markov chain (assuming the current state is ``theta_0`` and + that there are 2 proposal kernels):: + + theta_1 ~ N(theta_0, lambda * scale_1 * sigma) + alpha_1(theta_0, theta_1) = min(1, p(theta_1|X) / p(theta_0|X)) + u_1 ~ uniform(0, 1) + if alpha_1(theta_0, theta_1) > u_1: + return theta_1 + theta_2 ~ N(theta_0, lambda * scale_2 * sigma0) + alpha_2(theta_0, theta_1, theta_2) = + min(1, p(theta_2|X) (1 - alpha_1(theta_2, theta_1)) / + (p(theta_0|X) (1 - alpha_1(theta_0, theta_1)))) + u_2 ~ uniform(0, 1) + if alpha_2(theta_0, theta_1, theta_2) > u_2: + return theta_2 + else: + return theta_0 + + Our implementation also allows more than 2 proposal kernels to be used. + This means that ``k`` accept-reject steps are taken. In each step (``i``), + the probability that a proposal ``theta_i`` is accepted is:: + + alpha_i(theta_0, theta_1, ..., theta_i) = min(1, p(theta_i|X) / + p(theta_0|X) * n_i / d_i) + + where:: + + n_i = (1 - alpha_1(theta_i, theta_i-1)) * + (1 - alpha_2(theta_i, theta_i-1, theta_i-2)) * + ... + (1 - alpha_i-1(theta_i, theta_i-1, ..., theta_0)) + d_i = (1 - alpha_1(theta_0, theta_1)) * + (1 - alpha_2(theta_0, theta_1, theta_2)) * + ... + (1 - alpha_i-1(theta_0, theta_1, ..., theta_i-1)) + + If ``k`` proposals have been rejected, the initial point ``theta_0`` is + returned. + + At the end of each iterations, a 'base' proposal kernel is adapted:: + + mu = (1 - gamma) mu + gamma theta + sigma = (1 - gamma) sigma + gamma (theta - mu)(theta - mu)^t + log_lambda = log_lambda + gamma (accepted - target_acceptance_rate) + + where ``gamma = adaptations^-eta``, ``theta`` is the current state of + the Markov chain and ``accepted`` is a binary indicator for whether any of + the series of proposals were accepted. The kernels for the all proposals + are then adapted as ``[scale_1, scale_2, ..., scale_k] * sigma``, where the + scale factors are set using ``set_sigma_scale``. + + *Extends:* :class:`GlobalAdaptiveCovarianceMC` + + References + ---------- + .. [1] "DRAM: Efficient adaptive MCMC". + H Haario, M Laine, A Mira, E Saksman, (2006) Statistical Computing + https://doi.org/10.1007/s11222-006-9438-0 + """ + def __init__(self, x0, sigma0=None): + super(DramACMC, self).__init__(x0, sigma0) + + self._adapt_kernel = True + self._before_kernels_set = True + self._log_lambda = 0 + self._n_kernels = 2 + self._proposal_count = 0 + self._sigma_base = np.copy(self._sigma) + self._upper_scale = 1000 + self._Y = [None] * self._n_kernels + self._Y_log_pdf = np.zeros(self._n_kernels) + + def _adapt_sigma(self): + """ + Updates the covariance matrices of the various kernels being used + according to adaptive Metropolis routine. + """ + dsigm = np.reshape(self._current - self._mu, (self._n_parameters, 1)) + self._sigma_base = ((1 - self._gamma) * self._sigma_base + + self._gamma * np.dot(dsigm, dsigm.T)) + self._sigma = [self._sigma_scale[i] * self._sigma_base + for i in range(self._n_kernels)] + + def _calculate_alpha_log(self, n, Y, log_Y): + """ + Calculates alpha expression necessary in eq. 3 of Haario et al. for + determining accept/reject + """ + alpha_log = log_Y[n + 1] - log_Y[0] + if n == 0: + return min(0, alpha_log) + Y_rev = Y[::-1] + log_Y_rev = log_Y[::-1] + for i in range(n): + alpha_log += ( + stats.multivariate_normal.logpdf( + x=Y[n - i - 1], + mean=Y[n + 1], + cov=self._sigma[n], + allow_singular=True) - + stats.multivariate_normal.logpdf( + x=Y[i], + mean=self._current, + cov=self._sigma[0], + allow_singular=True) + + np.log(1 - np.exp(self._calculate_alpha_log( + i, Y_rev[0:(i + 2)], log_Y_rev[0:(i + 2)]))) - + np.log(1 - np.exp(self._calculate_alpha_log( + i, Y[0:(i + 2)], log_Y[0:(i + 2)]))) + ) + return min(0, alpha_log) + + def _calculate_r_log(self, fx): + """ + Calculates value of logged acceptance ratio (eq. 3 in [1]_). + """ + c = self._proposal_count + temp_Y = np.concatenate([[self._current], self._Y[0:(c + 1)]]) + temp_log_Y = np.concatenate( + [[self._current_log_pdf], self._Y_log_pdf[0:(c + 1)]]) + self._r_log = self._calculate_alpha_log(c, temp_Y, temp_log_Y) + + def _generate_proposal(self): + """ See :meth:`AdaptiveCovarianceMC._generate_proposal()`. """ + if self._before_kernels_set: + self.set_sigma_scale() + self._Y = [None] * self._n_kernels + self._Y_log_pdf = np.zeros(self._n_kernels) + self._before_kernels_set = False + + proposed = np.random.multivariate_normal( + self._current, np.exp(self._log_lambda) * + self._sigma[self._proposal_count]) + self._Y[self._proposal_count] = proposed + return proposed + + def name(self): + """ See :meth:`pints.MCMCSampler.name()`. """ + return 'Delayed Rejection Adaptive Metropolis (Dram) MCMC' + + def n_kernels(self): + """ Returns number of proposal kernels. """ + return self._n_kernels + + def n_hyper_parameters(self): + """ See :meth:`TunableMethod.n_hyper_parameters()`. """ + return 3 + + def set_hyper_parameters(self, x): + """ + The hyper-parameter vector is ``[eta, n_kernels, upper_scale]``. + + See :meth:`TunableMethod.set_hyper_parameters()`. + """ + self.set_eta(x[0]) + self.set_n_kernels(x[1]) + self.set_upper_scale(x[2]) + + def set_n_kernels(self, n_kernels): + """ Sets number of proposal kernels. """ + if n_kernels < 1: + raise ValueError('Number of proposal kernels must be equal to ' + + 'or greater than 1.') + self._n_kernels = int(n_kernels) + + def set_upper_scale(self, upper_scale): + """ + Set the upper scale of initial covariance matrix multipliers for each + of the kernels: ``[0,...,upper]`` where the gradations are uniform on + the log10 scale meaning the proposal covariance matrices are: + ``[10^upper,..., 1] * sigma``. + """ + if upper_scale < 0: + raise ValueError('Upper scale must be positive.') + self._upper_scale = upper_scale + + def set_sigma_scale(self): + """ + Set the scale of initial covariance matrix multipliers for each of the + kernels: ``[0,...,upper]`` where the gradations are uniform on the + log10 scale meaning the proposal covariance matrices are: + ``[10^upper,..., 1] * sigma``. + """ + a_min = np.log10(1) + a_max = np.log10(self._upper_scale) + self._sigma_scale = 10**np.linspace(a_min, a_max, self._n_kernels) + self._sigma_scale = self._sigma_scale[::-1] + self._sigma = [self._sigma_scale[i] * self._sigma_base + for i in range(self._n_kernels)] + + def sigma_scale(self): + """ + Returns scale factors used to multiply a base covariance matrix, + resulting in proposal matrices for each accept-reject step. + """ + return self._sigma_scale + + def tell(self, fx): + """ + If first proposal, then accept with ordinary Metropolis probability; if + a later proposal, use probability determined by [1]_. + """ + # Check if we had a proposal + if self._proposed is None: + raise RuntimeError('Tell called before proposal was set.') + + # Ensure fx is a float + fx = float(fx) + self._Y_log_pdf[self._proposal_count] = fx + + # First point? + if self._current is None: + if not np.isfinite(fx): + raise ValueError( + 'Initial point for MCMC must have finite logpdf.') + + # Accept + self._current = self._proposed + self._current_log_pdf = fx + + # Increase iteration count + self._iterations += 1 + + # Clear proposal + self._proposed = None + + # Return first point for chain + return self._current + + # Check if the proposed point can be accepted + accepted = 0 + + if np.isfinite(fx): + self._calculate_r_log(fx) + u = np.log(np.random.uniform(0, 1)) + if u < self._r_log: + accepted = 1 + self._current = self._proposed + self._current_log_pdf = fx + + self._proposed = None + + if accepted == 0: + # rejected proposal + if self._n_kernels > 1 and ( + self._proposal_count < (self._n_kernels - 1)): + self._proposal_count += 1 + return None + else: + self._proposal_count = 0 + self._gamma = (self._adaptations**-self._eta) + self._adaptations += 1 + + # Update mu, covariance matrix and log lambda + self._adapt_mu() + self._adapt_sigma() + self._log_lambda += (self._gamma * + (accepted - self._target_acceptance)) + return self._current + + def upper_scale(self): + """ + Returns upper scale limit (see + :meth:`pints.DramACMC.set_upper_scale()`). + """ + return self._upper_scale diff --git a/pints/tests/test_mcmc_dram_ac.py b/pints/tests/test_mcmc_dram_ac.py new file mode 100644 index 000000000..18e3eb475 --- /dev/null +++ b/pints/tests/test_mcmc_dram_ac.py @@ -0,0 +1,184 @@ +#!/usr/bin/env python +# +# Tests the basic methods of the Dram ACMC routine. +# +# This file is part of PINTS (https://github.com/pints-team/pints/) which is +# released under the BSD 3-clause license. See accompanying LICENSE.md for +# copyright notice and full license details. +# +import pints +import pints.toy as toy +import unittest +import numpy as np + +from shared import StreamCapture + +# Consistent unit testing in Python 2 and 3 +try: + unittest.TestCase.assertRaisesRegex +except AttributeError: + unittest.TestCase.assertRaisesRegex = unittest.TestCase.assertRaisesRegexp + + +class TestDramACMC(unittest.TestCase): + """ + Tests the basic methods of the adaptive covariance MCMC routine. + """ + + @classmethod + def setUpClass(cls): + """ Set up problem for tests. """ + + # Create toy model + cls.model = toy.LogisticModel() + cls.real_parameters = [0.015, 500] + cls.times = np.linspace(0, 1000, 1000) + cls.values = cls.model.simulate(cls.real_parameters, cls.times) + + # Add noise + cls.noise = 10 + cls.values += np.random.normal(0, cls.noise, cls.values.shape) + cls.real_parameters.append(cls.noise) + cls.real_parameters = np.array(cls.real_parameters) + + # Create an object with links to the model and time series + cls.problem = pints.SingleOutputProblem( + cls.model, cls.times, cls.values) + + # Create a uniform prior over both the parameters and the new noise + # variable + cls.log_prior = pints.UniformLogPrior( + [0.01, 400, cls.noise * 0.1], + [0.02, 600, cls.noise * 100] + ) + + # Create a log likelihood + cls.log_likelihood = pints.GaussianLogLikelihood(cls.problem) + + # Create an un-normalised log-posterior (log-likelihood + log-prior) + cls.log_posterior = pints.LogPosterior( + cls.log_likelihood, cls.log_prior) + + def test_method(self): + + # Create mcmc + x0 = self.real_parameters * 1.1 + mcmc = pints.DramACMC(x0) + + # Configure + mcmc.set_target_acceptance_rate(0.3) + mcmc.set_initial_phase(True) + + # Perform short run + rate = [] + chain = [] + for i in range(100): + x = mcmc.ask() + fx = self.log_posterior(x) + sample = mcmc.tell(fx) + while sample is None: + x = mcmc.ask() + fx = self.log_posterior(x) + sample = mcmc.tell(fx) + if i == 20: + mcmc.set_initial_phase(False) + if i >= 50: + chain.append(sample) + rate.append(mcmc.acceptance_rate()) + if np.all(sample == x): + self.assertEqual(mcmc.current_log_pdf(), fx) + + chain = np.array(chain) + rate = np.array(rate) + self.assertEqual(chain.shape[0], 50) + self.assertEqual(chain.shape[1], len(x0)) + self.assertEqual(rate.shape[0], 100) + + # Test with more kernels + x0 = self.real_parameters * 1.1 + mcmc = pints.DramACMC(x0) + + # Configure + mcmc.set_n_kernels(4) + + # Perform short run + rate = [] + chain = [] + for i in range(100): + x = mcmc.ask() + fx = self.log_posterior(x) + sample = mcmc.tell(fx) + while sample is None: + x = mcmc.ask() + fx = self.log_posterior(x) + sample = mcmc.tell(fx) + if i == 20: + mcmc.set_initial_phase(False) + if i >= 50: + chain.append(sample) + rate.append(mcmc.acceptance_rate()) + if np.all(sample == x): + self.assertEqual(mcmc.current_log_pdf(), fx) + + chain = np.array(chain) + rate = np.array(rate) + self.assertEqual(chain.shape[0], 50) + self.assertEqual(chain.shape[1], len(x0)) + self.assertEqual(rate.shape[0], 100) + + def test_options(self): + + # Test setting acceptance rate + x0 = self.real_parameters + mcmc = pints.DramACMC(x0) + self.assertRaises(RuntimeError, mcmc.tell, 0.0) + x0 = self.real_parameters + mcmc = pints.DramACMC(x0) + mcmc.ask() + self.assertRaises(ValueError, mcmc.tell, -float('inf')) + + self.assertNotEqual(mcmc.target_acceptance_rate(), 0.5) + mcmc.set_target_acceptance_rate(0.5) + self.assertEqual(mcmc.target_acceptance_rate(), 0.5) + mcmc.set_target_acceptance_rate(1) + self.assertRaises(ValueError, mcmc.set_target_acceptance_rate, 0) + self.assertRaises(ValueError, mcmc.set_target_acceptance_rate, -1e-6) + self.assertRaises(ValueError, mcmc.set_target_acceptance_rate, 1.00001) + + # test hyperparameter setters and getters + self.assertEqual(mcmc.n_hyper_parameters(), 3) + self.assertRaises(ValueError, mcmc.set_hyper_parameters, [-0.1, 1, 3]) + self.assertRaises(ValueError, mcmc.set_hyper_parameters, [0.5, 0, 3]) + self.assertRaises(ValueError, mcmc.set_hyper_parameters, [0.5, 1, -1]) + mcmc.set_hyper_parameters([0.1, 4, 3.5]) + self.assertEqual(mcmc.eta(), 0.1) + self.assertEqual(mcmc.n_kernels(), 4) + self.assertEqual(mcmc.upper_scale(), 3.5) + mcmc.ask() + mcmc.set_sigma_scale() + scale = mcmc.sigma_scale() + a_min = np.log10(1) + a_max = np.log10(3.5) + scale1 = 10**np.linspace(a_min, a_max, 4) + scale1 = scale1[::-1] + self.assertTrue(np.array_equal(scale, scale1)) + + self.assertEqual(mcmc.name(), ( + 'Delayed Rejection Adaptive Metropolis (Dram) MCMC')) + + def test_logging(self): + + # Test logging includes name. + x = [self.real_parameters] * 3 + mcmc = pints.MCMCController( + self.log_posterior, 3, x, method=pints.DramACMC) + mcmc.set_max_iterations(5) + with StreamCapture() as c: + mcmc.run() + text = c.text() + self.assertIn('Delayed Rejection Adaptive Metropolis (Dram) MCMC', + text) + + +if __name__ == '__main__': + unittest.main()