Model

class pinnx.nn.Model(input, approx, output, *args)

A neural network approximator.

Parameters:

• input (DictToArray) – The input check.

• approx (Module) – The neural network model.

• output (ArrayToDict) – The output unit.

gradient(inputs, order, y=None, *xi)

Compute the gradient of the approximator.

Parameters:

• inputs (Dict[str, Union[Array, ndarray, bool, number, bool, int, float, complex, Quantity]]) – The input data.

• order (int) – The order of the gradient.

• y (Union[str, Sequence[str], None]) – The output variables.

• xi (Union[str, Sequence[str], None]) – The input variables.

Returns:

The gradient of the approximator.

hessian(inputs, y=None, xi=None, xj=None)

Compute the Hessian of the approximator.

Compute: H[y][xi][xj] = d^2y / dxi dxj = d^2y / dxj dxi

Parameters:

• inputs (Dict[str, Union[Array, ndarray, bool, number, bool, int, float, complex, Quantity]]) – The input data.

• y (Union[str, Sequence[str], None]) – The output variables.

• xi (Union[str, Sequence[str], None]) – The first input variables.

• xj (Union[str, Sequence[str], None]) – The second input variables.

Returns:

The Hessian of the approximator.

jacobian(inputs, y=None, x=None)

Compute the Jacobian of the approximation neural networks.

Parameters:

• inputs (Dict[str, Union[Array, ndarray, bool, number, bool, int, float, complex, Quantity]]) – The input data.

• y (Union[str, Sequence[str], None]) – The output variables.

• x (Union[str, Sequence[str], None]) – The input variables.

Returns:

The Jacobian of the approximation neural networks.