Mathematical

formula.math module

Mathematical operations for the MRFM simulation.

mrfmsim.formula.math.as_strided_x(dataset, window)

Function for adjusting the stride size in the x direction.

The operation is very fast and does not require extra memory because it only defines the stride size rather than creating a new array For a dataset with a shape of (6, 5, 4) with a window of 3 in the x direction, the resulting array shape is (4, 5, 4). The two new parameters are (4, 6, 5, 4), stride is (160, 160, 32, 8) if each element is 2 bytes For example, to determine the max value of a 3-dimensional array for every 3 elements in x direction:

dataset_strided = strided_axis0(dataset, 3)
dataset_strided.max(axis = 1)

For more information, see as_strided and numpy arrays memory and strides.

Parameters:

• dataset (array) – the dataset target to determine max and min (or other running operations)

• window (int) – the size of a sliding window for the dataset

Returns:

strided dataset

Return type:

ndarray

mrfmsim.formula.math.slice_matrix(matrix, shape)

Slice numpy matrix.

The function only slices the matrix in the middle based on the shape. The resulting array should be a view of the original array, which provides memory and speed improvement.

Parameters:

• matrix (ndarray) – a numpy array.

• shape (tuple) – sliced shape, has the same dimension as the matrix. The shape along the sliced axis should be the same oddity as the matrix.