numojo.routines.math.differences

Difference routines for NuMojo (numojo.routines.math.differences).

Implements gradient and trapezoidal integration helpers for numerical differentiation and integration tasks.

Functions

gradient

gradient[dtype: DType = DType.float64](x: NDArray[dtype], spacing: Scalar[dtype]) -> NDArray[dtype]

Compute the gradient of y over x using the trapezoidal rule.

Constraints

fdtype must be a floating-point type if idtype is not a floating-point type.

Parameters:

• dtype (DType): Input data type.

Args:

• x (NDArray): An array.
• spacing (Scalar): An array of the same shape as x containing the spacing between adjacent elements.

Returns:

• NDArray

Raises

trapz

trapz[dtype: DType = DType.float64](y: NDArray[dtype], x: NDArray[dtype]) -> Scalar[dtype] where dtype.is_floating_point()

Compute the integral of y over x using the trapezoidal rule.

Constraints

x and y must have the same shape.

dtype must be a floating-point type.

Parameters:

• dtype (DType): The element type.

Args:

• y (NDArray): An array.
• x (NDArray): An array.

Returns:

• Scalar

Raises