Using custom arguments in boundary condition

[karthikncsuiisc]

I want to use custom arguments in my boundary conditions. I tried the below code and it gives error. This is a simple example. In the next step, I would like to use custom parameters along with input_ parameter in my boundary condition.

Is there a tutorial for implementing such a type?

class Eqn1D(TimeDependentProblem, SpatialProblem):
    def __init__(self,params):
        super(Eqn1D, self).__init__()

        self.params = params
        self.lbc = params.lbc

    # @classmethod
    # def get_params(cls):
    #     return cls.params

    # define the burger equation
    def ht_equation(input_, output_):
        du = grad(output_, input_)
        ddu = grad(du, input_, components=['dudx'])

        coef1 = params.alpha

        res = du.extract(['dudt'])-coef1*ddu.extract(['ddudxdx'])
        return res

    # define initial condition
    def initial_condition(input_, output_):
        u_expected = 0
        return output_.extract(['u']) - u_expected

    # assign output/ spatial and temporal variables
    output_variables = ['u']
    spatial_domain = CartesianDomain({'x': [0, 1]})
    temporal_domain = CartesianDomain({'t': [0, 1]})

    # problem condition statement
    conditions = {
        'gamma1': Condition(location=CartesianDomain({'x': 0, 't': [0, 1]}), equation=FixedGradient(-self.lbc,d='x')),
        'gamma2': Condition(location=CartesianDomain({'x': 1, 't': [0, 1]}), equation=FixedGradient(0,d='x')),
        't0': Condition(location=CartesianDomain({'x': [0, 1], 't': 0}), equation=Equation(initial_condition)),
        'D': Condition(location=CartesianDomain({'x': [0, 1], 't': [0, 1]}), equation=Equation(ht_equation)),
    }

[dario-coscia]

Hello 👋🏻 Please take a look at issue #225 , specifically this answer could help.
I would avoid defining an __init__ and use either static variables. The value of these variables can always be changed during training by:

# update
from pytorch_lightning.callbacks import Callback
class Update(Callback):
    def on_train_epoch_end(self, trainer, __):
        trainer.solver.problem.__class__.[STATIC VARIABLE NAME] = ...

For more on callbacks and where to put them have a look at the lightning documentation, it is just an extra argument of the Trainer!

Let me know how it goes and if you like the package leave us a star ⭐️ which helps us grow the community!😄

[karthikncsuiisc]

Thank you for the help this worked.