Intro

Constraint contains set of classes to impose conditions on the targets or their derivatives. This classes are designed as a way to impose constraints on different parts of targets and domain.

---

[source]

Data

sciann.constraints.data.Data(cond, name='data')

Data class to impose to the system.

Arguments

• cond: Functional. The Functional object that Data condition will be imposed on.
• name: String. A str for name of the pde.

Returns

Raises

• ValueError: 'cond' should be a functional object. 'mesh' should be a list of numpy arrays.

---

[source]

PDE

sciann.constraints.pde.PDE(pde, name='pde')

PDE class to impose to the system.

Arguments

• pde: Functional. The Functional object that pde if formed on.
• name: String. A str for name of the pde.

Returns

Raises

• ValueError: 'pde' should be a functional object.

---

[source]

Tie

sciann.constraints.tie.Tie(cond1, cond2, name='tie')

Tie class to constrain network outputs. constraint: cond1 - cond2 == sol.

Arguments

• cond1: Functional. A Functional object to be tied to cond2.
• cond2: Functional. A 'Functional' object to be tied to cond1.
• name: String. A str for name of the pde.

Returns

Raises

• ValueError: 'pde' should be a functional object.