2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 2017,2018,2019,2020, by the GROMACS development team, led by
5 * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
6 * and including many others, as listed in the AUTHORS file in the
7 * top-level source directory and at http://www.gromacs.org.
9 * GROMACS is free software; you can redistribute it and/or
10 * modify it under the terms of the GNU Lesser General Public License
11 * as published by the Free Software Foundation; either version 2.1
12 * of the License, or (at your option) any later version.
14 * GROMACS is distributed in the hope that it will be useful,
15 * but WITHOUT ANY WARRANTY; without even the implied warranty of
16 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
17 * Lesser General Public License for more details.
19 * You should have received a copy of the GNU Lesser General Public
20 * License along with GROMACS; if not, see
21 * http://www.gnu.org/licenses, or write to the Free Software Foundation,
22 * Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
24 * If you want to redistribute modifications to GROMACS, please
25 * consider that scientific software is very special. Version
26 * control is crucial - bugs must be traceable. We will be happy to
27 * consider code for inclusion in the official distribution, but
28 * derived work must not be called official GROMACS. Details are found
29 * in the README & COPYING files - if they are missing, get the
30 * official version at http://www.gromacs.org.
32 * To help us fund GROMACS development, we humbly ask that you cite
33 * the research papers on the package. Check out http://www.gromacs.org.
43 #include <gmock/gmock.h>
44 #include <gtest/gtest.h>
46 #include "gromacs/applied_forces/awh/bias.h"
47 #include "gromacs/applied_forces/awh/correlationgrid.h"
48 #include "gromacs/applied_forces/awh/pointstate.h"
49 #include "gromacs/mdtypes/awh_params.h"
50 #include "gromacs/utility/stringutil.h"
52 #include "testutils/refdata.h"
53 #include "testutils/testasserts.h"
61 //! The number of lambda states to use in the tests.
62 const int numLambdaStates = 16;
65 * Struct that gathers all input for setting up and using a Bias
67 struct AwhFepLambdaStateTestParameters
69 AwhFepLambdaStateTestParameters() = default;
71 AwhFepLambdaStateTestParameters(AwhFepLambdaStateTestParameters&& o) noexcept :
73 awhDimParams(o.awhDimParams),
74 awhBiasParams(o.awhBiasParams),
75 awhParams(o.awhParams),
76 dimParams(std::move(o.dimParams))
78 awhBiasParams.dimParams = &awhDimParams;
79 awhParams.awhBiasParams = &awhBiasParams;
81 double beta; //!< 1/(kB*T)
83 AwhDimParams awhDimParams; //!< Dimension parameters pointed to by \p awhBiasParams
84 AwhBiasParams awhBiasParams; //!< Bias parameters pointed to by \[ awhParams
85 AwhParams awhParams; //!< AWH parameters, this is the struct to actually use
87 std::vector<DimParams> dimParams; //!< Dimension parameters for setting up Bias
90 //! Helper function to set up the C-style AWH parameters for the test
91 static AwhFepLambdaStateTestParameters getAwhFepLambdaTestParameters(int eawhgrowth, int eawhpotential)
93 AwhFepLambdaStateTestParameters params;
97 AwhDimParams& awhDimParams = params.awhDimParams;
99 awhDimParams.period = 0;
100 awhDimParams.diffusion = 1e-4;
101 awhDimParams.origin = 0;
102 awhDimParams.end = numLambdaStates - 1;
103 awhDimParams.coordValueInit = awhDimParams.origin;
104 awhDimParams.coverDiameter = 0;
105 awhDimParams.eCoordProvider = eawhcoordproviderFREE_ENERGY_LAMBDA;
107 AwhBiasParams& awhBiasParams = params.awhBiasParams;
109 awhBiasParams.ndim = 1;
110 awhBiasParams.dimParams = &awhDimParams;
111 awhBiasParams.eTarget = eawhtargetCONSTANT;
112 awhBiasParams.targetBetaScaling = 0;
113 awhBiasParams.targetCutoff = 0;
114 awhBiasParams.eGrowth = eawhgrowth;
115 awhBiasParams.bUserData = FALSE;
116 awhBiasParams.errorInitial = 1.0 / params.beta;
117 awhBiasParams.shareGroup = 0;
118 awhBiasParams.equilibrateHistogram = FALSE;
120 int64_t seed = 93471803;
122 params.dimParams.push_back(DimParams::fepLambdaDimParams(numLambdaStates, params.beta));
124 AwhParams& awhParams = params.awhParams;
126 awhParams.numBias = 1;
127 awhParams.awhBiasParams = &awhBiasParams;
128 awhParams.seed = seed;
129 awhParams.nstOut = 0;
130 awhParams.nstSampleCoord = 1;
131 awhParams.numSamplesUpdateFreeEnergy = 10;
132 awhParams.ePotential = eawhpotential;
133 awhParams.shareBiasMultisim = FALSE;
138 //! Convenience typedef: growth type enum, potential type enum, disable update skips
139 typedef std::tuple<int, int, BiasParams::DisableUpdateSkips> BiasTestParameters;
141 /*! \brief Test fixture for testing Bias updates
143 class BiasFepLambdaStateTest : public ::testing::TestWithParam<BiasTestParameters>
146 //! Random seed for AWH MC sampling
150 std::unique_ptr<Bias> bias_;
152 BiasFepLambdaStateTest()
154 /* We test all combinations of:
156 * eawhgrowthLINEAR: final, normal update phase
157 * ewahgrowthEXP_LINEAR: intial phase, updated size is constant
158 * eawhpotential (test both, but for the FEP lambda state dimension MC will in practice be used,
159 * except that eawhpotentialCONVOLVED also gives a potential output):
160 * eawhpotentialUMBRELLA: MC on lambda state
161 * eawhpotentialCONVOLVED: MD on a convolved potential landscape (falling back to MC on lambda state)
162 * disableUpdateSkips (should not affect the results):
163 * BiasParams::DisableUpdateSkips::yes: update the point state for every sample
164 * BiasParams::DisableUpdateSkips::no: update the point state at an interval > 1 sample
166 * Note: It would be nice to explicitly check that eawhpotential
167 * and disableUpdateSkips do not affect the point state.
168 * But the reference data will also ensure this.
172 BiasParams::DisableUpdateSkips disableUpdateSkips;
173 std::tie(eawhgrowth, eawhpotential, disableUpdateSkips) = GetParam();
175 /* Set up a basic AWH setup with a single, 1D bias with parameters
176 * such that we can measure the effects of different parameters.
178 const AwhFepLambdaStateTestParameters params =
179 getAwhFepLambdaTestParameters(eawhgrowth, eawhpotential);
181 seed_ = params.awhParams.seed;
183 double mdTimeStep = 0.1;
185 bias_ = std::make_unique<Bias>(-1, params.awhParams, params.awhBiasParams, params.dimParams,
186 params.beta, mdTimeStep, 1, "", Bias::ThisRankWillDoIO::No,
191 TEST_P(BiasFepLambdaStateTest, ForcesBiasPmf)
193 gmx::test::TestReferenceData data;
194 gmx::test::TestReferenceChecker checker(data.rootChecker());
198 /* Make strings with the properties we expect to be different in the tests.
199 * These also helps to interpret the reference data.
201 std::vector<std::string> props;
202 props.push_back(formatString("stage: %s", bias.state().inInitialStage() ? "initial" : "final"));
203 props.push_back(formatString("convolve forces: %s", bias.params().convolveForce ? "yes" : "no"));
204 props.push_back(formatString("skip updates: %s", bias.params().skipUpdates() ? "yes" : "no"));
206 SCOPED_TRACE(gmx::formatString("%s, %s, %s", props[0].c_str(), props[1].c_str(), props[2].c_str()));
208 std::vector<double> force, pot;
210 double potentialJump = 0;
211 double mdTimeStep = 0.1;
214 /* Some energies to use as base values (to which some noise is added later on). */
215 std::vector<double> neighborLambdaEnergies(numLambdaStates);
216 std::vector<double> neighborLambdaDhdl(numLambdaStates);
217 const double magnitude = 12.0;
218 for (int i = 0; i < numLambdaStates; i++)
220 neighborLambdaEnergies[i] = magnitude * std::sin(i * 0.1);
221 neighborLambdaDhdl[i] = magnitude * std::cos(i * 0.1);
224 for (int step = 0; step < nSteps; step++)
226 int umbrellaGridpointIndex = bias.state().coordState().umbrellaGridpoint();
227 awh_dvec coordValue = { bias.getGridCoordValue(umbrellaGridpointIndex)[0], 0, 0, 0 };
228 double potential = 0;
229 gmx::ArrayRef<const double> biasForce = bias.calcForceAndUpdateBias(
230 coordValue, neighborLambdaEnergies, neighborLambdaDhdl, &potential, &potentialJump,
231 nullptr, nullptr, step * mdTimeStep, step, seed_, nullptr);
233 force.push_back(biasForce[0]);
234 pot.push_back(potential);
237 /* When skipping updates, ensure all skipped updates are performed here.
238 * This should result in the same bias state as at output in a normal run.
240 if (bias.params().skipUpdates())
242 bias.doSkippedUpdatesForAllPoints();
245 std::vector<double> pointBias, logPmfsum;
246 for (auto& point : bias.state().points())
248 pointBias.push_back(point.bias());
249 logPmfsum.push_back(point.logPmfSum());
252 constexpr int ulpTol = 10;
254 checker.checkSequence(props.begin(), props.end(), "Properties");
255 checker.setDefaultTolerance(absoluteTolerance(magnitude * GMX_DOUBLE_EPS * ulpTol));
256 checker.checkSequence(force.begin(), force.end(), "Force");
257 checker.checkSequence(pot.begin(), pot.end(), "Potential");
258 checker.setDefaultTolerance(relativeToleranceAsUlp(1.0, ulpTol));
259 checker.checkSequence(pointBias.begin(), pointBias.end(), "PointBias");
260 checker.checkSequence(logPmfsum.begin(), logPmfsum.end(), "PointLogPmfsum");
263 /* Scan initial/final phase, MC/convolved force and update skip (not) allowed
264 * Both the convolving and skipping should not affect the bias and PMF.
265 * It would be nice if the test would explicitly check for this.
266 * Currently this is tested through identical reference data.
268 INSTANTIATE_TEST_CASE_P(WithParameters,
269 BiasFepLambdaStateTest,
270 ::testing::Combine(::testing::Values(eawhgrowthLINEAR, eawhgrowthEXP_LINEAR),
271 ::testing::Values(eawhpotentialUMBRELLA, eawhpotentialCONVOLVED),
272 ::testing::Values(BiasParams::DisableUpdateSkips::yes,
273 BiasParams::DisableUpdateSkips::no)));
275 // Test that we detect coverings and exit the initial stage at the correct step
276 TEST(BiasFepLambdaStateTest, DetectsCovering)
278 const AwhFepLambdaStateTestParameters params =
279 getAwhFepLambdaTestParameters(eawhgrowthEXP_LINEAR, eawhpotentialCONVOLVED);
281 const double mdTimeStep = 0.1;
283 Bias bias(-1, params.awhParams, params.awhBiasParams, params.dimParams, params.beta, mdTimeStep,
284 1, "", Bias::ThisRankWillDoIO::No);
286 const int64_t exitStepRef = 320;
288 bool inInitialStage = bias.state().inInitialStage();
290 /* Some energies to use as base values (to which some noise is added later on). */
291 std::vector<double> neighborLambdaEnergies(numLambdaStates);
292 std::vector<double> neighborLambdaDhdl(numLambdaStates);
293 const double magnitude = 12.0;
294 for (int i = 0; i < numLambdaStates; i++)
296 neighborLambdaEnergies[i] = magnitude * std::sin(i * 0.1);
297 neighborLambdaDhdl[i] = magnitude * std::cos(i * 0.1);
301 /* Normally this loop exits at exitStepRef, but we extend with failure */
302 for (step = 0; step <= 2 * exitStepRef; step++)
304 int umbrellaGridpointIndex = bias.state().coordState().umbrellaGridpoint();
305 awh_dvec coordValue = { bias.getGridCoordValue(umbrellaGridpointIndex)[0], 0, 0, 0 };
307 double potential = 0;
308 double potentialJump = 0;
309 bias.calcForceAndUpdateBias(coordValue, neighborLambdaEnergies, neighborLambdaDhdl,
310 &potential, &potentialJump, nullptr, nullptr, step, step,
311 params.awhParams.seed, nullptr);
313 inInitialStage = bias.state().inInitialStage();
320 EXPECT_EQ(false, inInitialStage);
323 EXPECT_EQ(exitStepRef, step);