Fix bug with calculations on small timescales

[KulaginVladimir]

Proposed changes

When performing simulations on small timescales (e.g., final_time=1e-7), some bugs appear due to the use of np.isclose with the default parameters (atol, rtol). For example (see below), computation stops without reaching the final time. As @RemDelaporteMathurin suggested, the current PR attempts to fix the bug by setting atol=0 in np.isclose.

Types of changes

What types of changes does your code introduce to FESTIM?

• Bugfix (non-breaking change which fixes an issue)
• New feature (non-breaking change which adds functionality)
• Breaking change (fix or feature that would cause existing functionality to not work as expected)
• Code refactoring
• Documentation Update (if none of the other choices apply)
• New tests

Checklist

• Black formatted
• Unit tests pass locally with my changes
• I have added tests that prove my fix is effective or that my feature works
• I have added necessary documentation (if appropriate)

Further comments

MWE:

import festim as F
import numpy as np

my_model = F.Simulation(log_level = 40)

my_model.mesh = F.MeshFromVertices(np.linspace(0, 1, 10))

my_model.materials = F.Material(
    id=1,
    D_0=1,
    E_D=0.1,
)

my_model.T = F.Temperature(1000)

my_model.settings = F.Settings(
    absolute_tolerance=1e-10,
    relative_tolerance=1e-10,
    final_time=1e-7,
)

my_model.dt = F.Stepsize(
    initial_value=1e-10,
)

my_model.initialise()
my_model.run()

Stops at 90%:

Defining initial values
Defining variational problem
Defining source terms
Defining boundary conditions
Time stepping...
90.0 %        9.0e-08 s    Ellapsed time so far: 3.2 s

[RemDelaporteMathurin]

Very nice MWE @KulaginVladimir , would you mind making use of it in a test that catches the bug?

I think you can do something like:

def test_small_timesteps_final_time_bug():
    """Some description"""
    my_model = F.Simulation(log_level = 40)
    
    my_model.mesh = F.MeshFromVertices(np.linspace(0, 1, 10))
    
    my_model.materials = F.Material(
        id=1,
        D_0=1,
        E_D=0.1,
    )
    
    my_model.T = F.Temperature(1000)
    
    my_model.settings = F.Settings(
        absolute_tolerance=1e-10,
        relative_tolerance=1e-10,
        final_time=1e-7,
    )
    
    my_model.dt = F.Stepsize(
        initial_value=1e-10,
    )
    
    my_model.initialise()
    my_model.run()

    assert np.isclose(my_model.t, my_model.settings.final_time, atol=0.0)

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (7c6ca17) 98.93% compared to head (0d3dd1e) 98.93%.
Report is 3 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main     #682   +/-   ##
=======================================
  Coverage   98.93%   98.93%           
=======================================
  Files          57       57           
  Lines        2154     2154           
=======================================
  Hits         2131     2131           
  Misses         23       23           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[RemDelaporteMathurin]

Hi @KulaginVladimir just a small comment on my side, Thanks for this fix! 🚀

[RemDelaporteMathurin]

Looks good to me thanks @KulaginVladimir