Update Physical Constants and Corresponding Unit Tests #2866
Conversation
…ests that this change introduced
|
This is the current status of the tests: =========================== short test summary info ============================
FAILED test/arkane/arkaneOutputTest.py::OutputTest::test_prettify - assert " E0 = (193.749, 'kJ/mol'),\n" in ['# Coordinates for C7H7 in Input Orientation (angstroms):\n', '# C 0.0000 0.0000 -1.9093\n', '# C 0.0000 1.2097 -1.2046\n', '# C 0.0000 1.2164 0.1782\n', '# C 0.0000 0.0000 0.9209\n', '# C 0.0000 -1.2164 0.1782\n', '# C 0.0000 -1.2097 -1.2046\n', '# H 0.0000 0.0000 -2.9930\n', '# H 0.0000 2.1490 -1.7469\n', '# H 0.0000 2.1576 0.7177\n', '# H 0.0000 -2.1576 0.7177\n', '# H 0.0000 -2.1490 -1.7469\n', '# C 0.0000 0.0000 2.3249\n', '# H 0.0000 0.9273 2.8840\n', '# H 0.0000 -0.9273 2.8840\n', 'conformer(\n', " label = 'C7H7',\n", " E0 = (193.688, 'kJ/mol'),\n", ' modes = [\n', " IdealGasTranslation(mass=(91.0548, 'amu')),\n", ' NonlinearRotor(\n', " inertia = ([91.0567, 186.675, 277.733], 'amu*angstrom^2'),\n", ' symmetry = 2,\n', ' ),\n', ' HarmonicOscillator(\n', " frequencies = ([199.381, 360.536, 413.795, 480.347, 536.285, 630.723, 687.118, 709.613, 776.662, 830.404, 834.386, 901.841, 973.498, 975.148, 993.349, 998.606, 1040.14, 1120.69, 1179.22, 1189.07, 1292.86, 1332.91, 1357.18, 1479.46, 1495.36, 1507.91, 1583.14, 1604.63, 3156.85, 3170.22, 3172.78, 3185.05, 3189.8, 3203.23, 3253.99], 'cm^-1'),\n", ' ),\n', ' HinderedRotor(\n', " inertia = (1.70013, 'amu*angstrom^2'),\n", ' symmetry = 2,\n', ' fourier = (\n', ' [\n', ' [-0.31592, -27.7665, 0.177659, 3.24367, 0.0509783],\n', ' [-0.00152732, -0.00192155, -0.00354267, -0.000523773, 0.00324541],\n', ' ],\n', " 'kJ/mol',\n", ' ),\n', ' quantum = None,\n', ' semiclassical = None,\n', ' ),\n', ' ],\n', ' spin_multiplicity = 2,\n', ' optical_isomers = 1,\n', ')\n', '\n', '# Thermodynamics for C7H7:\n', '# Enthalpy of formation (298 K) = 50.553 kcal/mol\n', '# Entropy of formation (298 K) = 75.505 cal/(mol*K)\n', '# =========== =========== =========== =========== ===========\n', '# Temperature Heat cap. Enthalpy Entropy Free energy\n', '# (K) (cal/mol*K) (kcal/mol) (cal/mol*K) (kcal/mol)\n', '# =========== =========== =========== =========== ===========\n', '# 300 25.640 50.604 75.676 27.902\n', '# 400 33.155 53.551 84.101 19.911\n', '# 500 39.702 57.202 92.222 11.092\n', '# 600 45.261 61.458 99.966 1.479\n', '# 800 53.880 71.423 114.239 -19.969\n', '# 1000 59.876 82.836 126.948 -44.112\n', '# 1500 67.906 115.082 152.988 -114.400\n', '# 2000 71.822 150.084 173.100 -196.117\n', '# 2400 74.219 179.310 186.415 -268.086\n', '# =========== =========== =========== =========== ===========\n', 'thermo(\n', " label = 'C7H7',\n", ' thermo = NASA(\n', ' polynomials = [\n', ' NASAPolynomial(\n', ' coeffs = [3.95186, 0.0029882, 0.000162149, -2.93429e-07, 1.70812e-10, 23197, 9.643],\n', " Tmin = (10, 'K'),\n", " Tmax = (445.411, 'K'),\n", ' ),\n', ' NASAPolynomial(\n', ' coeffs = [-2.82364, 0.0638647, -4.29617e-05, 1.37173e-08, -1.66592e-12, 23800.3, 36.8479],\n', " Tmin = (445.411, 'K'),\n", " Tmax = (3000, 'K'),\n", ' ),\n', ' ],\n', " Tmin = (10, 'K'),\n", " Tmax = (3000, 'K'),\n", " E0 = (192.856, 'kJ/mol'),\n", " Cp0 = (33.2579, 'J/(mol*K)'),\n", " CpInf = (328.421, 'J/(mol*K)'),\n", ' ),\n', ')\n', '\n']
FAILED test/rmgpy/reactionTest.py::TestChargeTransferReaction::test_get_rate_coeff - assert 0.0078125 < 1e-06
+ where 0.0078125 = abs((43870307169260.06 - 43870307169260.055))
== 2 failed, 1968 passed, 39 skipped, 3937806 warnings in 2329.92s (0:38:49) ===These both look like regular old rounding errors. The first one is because Thanks again @ttpekkan! |
JacksonBurns
changed the title
Regression Testing Results
Detailed regression test results.Regression test aromatics:Reference: Execution time (DD:HH:MM:SS): 00:00:00:49 aromatics Passed Core Comparison ✅Original model has 15 species. aromatics Passed Edge Comparison ✅Original model has 106 species. DetailsObservables Test Case: Aromatics Comparison✅ All Observables varied by less than 0.500 on average between old model and new model in all conditions! aromatics Passed Observable Testing ✅Regression test liquid_oxidation:Reference: Execution time (DD:HH:MM:SS): 00:00:01:56 liquid_oxidation Passed Core Comparison ✅Original model has 37 species. liquid_oxidation Failed Edge Comparison ❌Original model has 214 species. Non-identical kinetics! ❌
kinetics: DetailsObservables Test Case: liquid_oxidation Comparison✅ All Observables varied by less than 0.100 on average between old model and new model in all conditions! liquid_oxidation Passed Observable Testing ✅Regression test nitrogen:Reference: Execution time (DD:HH:MM:SS): 00:00:01:02 nitrogen Passed Core Comparison ✅Original model has 41 species. nitrogen Passed Edge Comparison ✅Original model has 133 species. DetailsObservables Test Case: NC Comparison✅ All Observables varied by less than 0.200 on average between old model and new model in all conditions! nitrogen Passed Observable Testing ✅Regression test oxidation:Reference: Execution time (DD:HH:MM:SS): 00:00:01:48 oxidation Passed Core Comparison ✅Original model has 59 species. oxidation Passed Edge Comparison ✅Original model has 230 species. DetailsObservables Test Case: Oxidation Comparison✅ All Observables varied by less than 0.500 on average between old model and new model in all conditions! oxidation Passed Observable Testing ✅Errors occurred during observable testing
WARNING:root:Initial mole fractions do not sum to one; normalizing.
|
2 participants
Motivation or Problem
RMG currently uses outdated values for many physical constants. Many of these constants were given exact values when they were last updated, so now we have an opportunity to fix the values of these once and for all. This likely will have little to no difference in practice but is the correct thing to do.
Description of Changes
I changed the physical constants to the values reported here: https://arxiv.org/pdf/2409.03787 For atomic weights, I used the most recent values I found from IUPAC: https://iupac.qmul.ac.uk/AtWt/ Note that for some important elements like H and O it's not possible to give an unambiguous isotope-corrected atomic weight as it depends on the environment. I just went with the average value given in the table. Also, converting the dimensionless atomic weights to mol/kg requires multiplication by the molar mass constant (1.0000000010531e-3), which is no longer exactly 1.0e-3 due to how the physical constant were updated. I did this conversion, and although the correct thing to do, does not make a difference in practice. This also slightly affects the atomic mass unit as it is defined as the molar mass constant divided by Avogadro's constant. I did not include unstable elements. These can be added as needed, but I don't think RMG currently supports using them anyway.
Testing
I ran make–test, and as expected, it threw quite a few errors because in the unit tests some RMG output values are compared with pre-defined values. The RMG output values changed a tiny amount because of the physical-constant and atomic-weight updates, but enough for the unit tests to fail. I simply updated the pre-set values in the unit tests to those that RMG outputs with the updated physical constant and atomic weights. In a few cases I adjusted the tolerances for approximate equality comparisons to make the unit tests pass. Most of the errors appear to come from the change to the Hartree --> kJ/mol conversion factor.
Reviewer Tips
I guess what mostly needs to be done is to check that the physical constants do not have typos.