Hi Ning,

Thanks for your kind words. That part of the lesson is to rationalize the outcome. So instead of having 250 different sized squares, we decide we only want 5(or 6 or 10) different sizes. There are a lot of different ways of doing this and using the replace paths + find domain is one of them.

Let's start with the replace paths component. The only thing this component does is reorganize a data tree or list. The Param viewer is just to get the path number {0;0}, {0;1} etc. from the grafted input and use as a search mask. We could also use an asterisk as we want to replace all the inputs, for some reason however Grasshopper throws an error. Using the param viewer to get all the path names/numbers for the search mask is a workaround.

Below is an image explaining the Find domain, I hope this is more clear.

The step of rationalizing is not included, but I explain this in this course: https://thinkparametric.com/courses/panel-clustering-using-machine-learning-soumaya-museum

It includes replacing the squares in a group for one size square.

Best,

Arie