simpler_stride_ticks.py

import numpy as np
from numpy.lib.stride_tricks import as_strided

def broadcast_to(array, shape):
    """Expand a numpy.ndarray to a new shape according to broadcasting rules
    """
    array = np.asarray(array)
   
    # will raise ValueError if shapes incompatible 
    np.nditer((array,), itershape=shape)

    strides = ([0] * (len(shape) - array.ndim)
               + [0 if size == 1 else stride
                  for size, stride
                  in zip(array.shape, array.strides)])
    return as_strided(array, shape=shape, strides=strides)

def broadcast_shape(*args):
    return np.nditer(args, flags=['multi_index']).shape 

def broadcast_arrays(*args):
    shape = broadcast_shape(*args)
    return [broadcast_to(array, shape) for array in args]

I think broadcast_to and broadcast_arrays could also be a one-liners:

def broadcast_to(array, shape)
    return np.nditer((array,), flags=['multi_index'], itershape=shape).itviews[0]

def broadcast_arrays(*args):
    return np.nditer(args, flags=['multi_index']).itviews