Parameters and Formulas

The parameter–formula system lets you link model quantities to named parameters, making parametric sweeps and sensitivity studies straightforward.

Important

The parameter–formula system is in active development and subject to change. The API described here is not stable and will likely change in future releases.

Concepts

Parameters

A named key–value store. Values can be changed between solver calls.

Entities

A handle that points to one scalar quantity in the model — a node attribute or a coupling value.

Formulas

A rule that writes a parameter’s value into an entity when apply_formulas() is called.

Important

Formulas are not auto-applied. You must call tmm.formulas.apply_formulas() to propagate parameter changes to the model before solving.

Creating entities

from pycanha.parameters import Entity

# Node 1's internal dissipation
qi_entity = Entity.qi(tmm.network, 1)

# Conductive coupling GL(1, 2)
gl_entity = Entity.gl(tmm.network, 1, 2)

# Radiative coupling GR(2, 3)
gr_entity = Entity.gr(tmm.network, 2, 3)

Each entity exposes get_value() / set_value() to read or write the underlying model quantity directly, and string_representation() for a human-readable label (e.g. "QI1", "GL(1, 2)").

Linking parameters to entities

from pycanha.parameters import ParameterFormula

# Register a parameter
tmm.parameters.add_parameter("k", 1.0)

# Create the formula: GL(1,2) = k
formula = ParameterFormula(gl_entity, tmm.parameters, "k")
tmm.formulas.add_formula(formula)

Running a parameter analysis

import numpy as np
from pycanha.solvers import SSLU

solver = SSLU(tmm)
solver.initialize()

k_values = np.linspace(0.5, 10.0, 50)
temperatures = []

for k in k_values:
    tmm.parameters.set_parameter("k", k)
    tmm.formulas.apply_formulas()      # propagate k → GL(1,2)
    solver.solve()
    temperatures.append(tmm.nodes.get_T(1))

solver.deinitialize()

ValueFormula

ValueFormula stores a fixed value inside the formula object. It is useful for freezing an entity at a constant that can be changed programmatically:

from pycanha.parameters import ValueFormula

vf = ValueFormula(qi_entity)
vf.set_value(42.0)
tmm.formulas.add_formula(vf)
tmm.formulas.apply_formulas()
# QI1 is now 42.0