Tests missing

[erjank]

What properties should our simulation have, and how can we test that we're passing these minimal sanity checks as we develop?

[erjank]

Maybe one test is that a regular MC simulation of the densest disk packing should have an acceptance ratio of 0? Or for two LJ particles the average PE should be close to -1.0?

[marjanalbooyeh]

Running the test for a system with 2 LJ particles using below parameters:

sim = Simulation(n_particles=2, n_density=0.1, 
                 energy_func=lj_energy, 
                 trajectory_write_freq=10, energy_write_freq=10, hard_sphere=False)
sim.run(500000, kT=2.0, max_trans=0.01)

Results:

• sim.acceptance_ratio: 0.002848
• Average energy: -0.99993

[erjank]

I'm a little surprised with a kT of 2 that there aren't more fluctuations in energy. A test with a triangle of 3 particles would be another good test (a) b/c each particle has 2 neighbors (which is a lot more complicated than each particle just having 1), and (b) b/c it's easy to calculate the average energy of -3 at low temperature.

…

On Tue, Jan 31, 2023 at 9:06 AM Marjan Albooyeh ***@***.***> wrote: Running the test for a system with 2 LJ particles using below parameters: sim = Simulation(n_particles=2, n_density=0.1, energy_func=lj_energy, trajectory_write_freq=10, energy_write_freq=10, hard_sphere=False) sim.run(500000, kT=2.0, max_trans=0.01) *Results:* - sim.acceptance_ratio: 0.002848 - Average energy: -0.99993 [image: image] <https://user-images.githubusercontent.com/16212171/215814661-7d748e35-3ed9-456d-bde5-69ce47136fb6.png> — Reply to this email directly, view it on GitHub <#11 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABIJRSWCTDSBPAPGIUV3KP3WVE2B3ANCNFSM6AAAAAAT2WHBNM> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[marjanalbooyeh]

Results from running the simulation for 3 LJ particles:

sim = Simulation(n_particles=3, n_density=0.1, 
                 energy_func=lj_energy, 
                 trajectory_write_freq=10000, energy_write_freq=10000, hard_sphere=False)
sim.run(50000000, kT=1.0, max_trans=0.05)

Results:

• sim.acceptance_ratio : 0.1811
• Average energy: 24310669.286873266
• decorrelation steps: 1
  

Since it seems like the system is not converging yet (high average energy and decorrelation steps is 1), I took a look at the histogram of the energy and it's seems like energy is concentrated around -3.

Does that make sense @erjank?

[erjank]

I'm guessing this code isn't any different from the code running the 2LJ case, right? We've got tons of "overlap" here. At kT=1, I would expect no energies above 0, and most energies around - 3 The same problems we mentioned in the 2LJ case are here in the 3LJ case - something is wrong in the energy calculation or the acceptance criteria, or both

…

On Tue, Jan 31, 2023 at 11:15 AM Marjan Albooyeh ***@***.***> wrote: Results from running the simulation for 3 LJ particles: sim = Simulation(n_particles=3, n_density=0.1, energy_func=lj_energy, trajectory_write_freq=10000, energy_write_freq=10000, hard_sphere=False) sim.run(50000000, kT=1.0, max_trans=0.05) *Results:* - sim.acceptance_ratio : 0.1811 - Average energy: 24310669.286873266 - decorrelation steps: 1 [image: image] <https://user-images.githubusercontent.com/16212171/215847042-a350f47b-7d16-45bc-88c7-53d91a0ffaf2.png> Since it seems like the system is not converging yet (high average energy and decorrelation steps is 1), I took a look at the histogram of the energy and it's seems like energy is concentrated around -3. [image: Screen Shot 2023-01-31 at 11 13 54 AM] <https://user-images.githubusercontent.com/16212171/215847449-6b87c544-c8c6-47d5-ba10-52462498957b.png> Does that make sense @erjank <https://github.com/erjank>? — Reply to this email directly, view it on GitHub <#11 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABIJRSV2D4KWJA3TPVURNLDWVFJCVANCNFSM6AAAAAAT2WHBNM> . You are receiving this because you were mentioned.Message ID: ***@***.***>