I have a simple graph with the attributes height and area

I have a photo (coverted into an array) which contains x,y and a f(x,y) parameter for each x and y. Effectivley I want to simulate random points in the photo and then use a minimum cost algrithm to move between the points by minimising the penalties given by the f(x,y). I have looked into creating a network (graph using NetworkX) but keep getting memory errors because the graph ends up being so big. I’m open to other ideas other than NetworkX if you think theres better ways to solve my problem.

I’m trying to use graphblas algorithms to speed up computation on a big network using Python. Unfortunately, the results returned from graphblas are different from the ones given by network algorithms even for very trivial networks. Here is a minimal working example: