I’m making a game where there are many Area2Ds of different countries, but I want to know the size of each country to calculate their surrender limit (bigger countries take longer to surrender, for example).

I don’t want to find the area of just one collision2D as some countries have multiple (see island nations, overseas colonies…), but looping that with the children could work if finding area using AREA2ds isn’t a thing.

I don’t want to simply use the outermost points of a collisionshape2D and draw a big rectangle, as it would make some countries way bigger than they really are (see especially the British and French empires)

Is it possible to find the area of an Area2D? If not, is it possible to calculate the area of CollisionShape2Ds (then I could just loop through the children).

p.s. Country size will not be the only factor for surrender limit, things like equipment stockpiles and something similar to war support in HOI4 might also affect it…

  • Adalast@lemmy.world
    2·
    15 days ago

    You mentioned the areas being countries. This leads me to believe that they are concave figures, correct? If you are unfamiliar, a concave figure is something that has a space that doubles back into the interior of the shape. So an o is convex, and a c is concave, as an example. Convex shapes are much simpler to find the area of. You can use a Riemann Sum as others have suggested. I would probably just pick a point inside the shape and do a bunch of triangles around with the point as an apex and the bases as two points on the edge of the surface, then sum up the areas of each triangle. You could even probably use a triangulation algorithm built into the engine to do this. (I am unfamiliar with the specifics of the Godot engine).

    For concave shapes it becomes a little more complex. It has been mentioned that you can draw a bounding box around the shape, so that would allow you to calculate it using a numerical method. Take random samples inside the bounding box and count up the number that are inside the shape and divide by the total number of samples. The value you get will be the % of the area of the bounding rectangle that the shape takes up, so just multiply the easy area by the % and you will get an answer that is close enough. It may take a bit to get the sample count right, but it will get there. Try to make sure the samples are as uniform as possible. You could even scatter the sample points then relax them for a couple iterations before counting to increase accuracy without increasing samples.