Data distribution in a Tree

Replies to This Discussion

Permalink Reply by pBHATT on June 4, 2017 at 5:26am

Basically what I am trying to find out is why is that {0;0} branch needed. If we assign a point to the point container, no branch s created and the data is stored in the {0} itself, so why then in grid component it starts from {0;0}.

Permalink Reply by David Rutten on June 4, 2017 at 5:59am

{0} and {0;0} do not contain anything, they don't in fact exist. Their existence is implied by the general structure of the tree. If you double-click the Param Viewer, you'll see the straight up list of branches that do exist, and nothing more.

The depth of the branching is determined by an algorithm which decides how much more complex the data becomes passing through the component. If all the inputs are [item] level and all outputs are [item] level, then the complexity is retained. If all inputs are [list] level and all outputs are [list] level, then the complexity is retained. If, however, some inputs are sometimes [item] level and some outputs are [list] level, then an extra layer of complexity it added to the tree branches. The reason for this is that we want the accumulation of complexity to be consistent. The output of a component should not become qualitatively different if you input one value vs. inputting two values. We only want quantitative differences.

Permalink Reply by pBHATT on June 4, 2017 at 9:53am

Hello David,

Thanks for the help :)