I have been using decision trees and discovered there is a core principle in all of this that extends to all data. I will blog about it when I have tested to make certain, but it seems to be the ultimate AI method that solves all relationships. It may become clear in an example I am creating in OpenGL with a C module that does the trees. It reflects the decision tree back upon itself. It may be difficult to envision without a concrete example of how it is applied to achieve something that many spend decades analyzing, while achieving only moderate success. I am fairly certain that it is correct, but I will test with real data.
The python example code and image it generates is just some part of a side investigation that prompted the discovery. It is actually the Big Dipper, which is of course not a dipper at all when viewed at most other angles. I was actually looking for the data base that Stellarium uses, so I could devise a sensible positioning mechanism for applying a technique similar to Star Trek long range scan. I had not considered its application in near space, but it would seem that I could make a long range scan of everything within a light hour of the current position. It might reveal some interesting things, if it works well. I have no doubt it will work as I have tested it. The real question is how much resolution is possible.
This is a second part. It is not so much the content as the association of content type, date, format and information which created the vector association, like a note heard in a symphony which implies a relationship of parts that becomes a vector set in normalized n-dimensional vector phase space. To put it simply, it was "like" something else. And that was a solution to a problem that I only had the rough shape of the empty (NULL) space to state it more mathematically.
The image left is part of a NASA document from 1967. I would not have been able to have access then or perhaps would have looked at it in a different way from the vector of my understanding at the time. The combination of certainty and possibility is never exhausted and even with a mechanical or electrical method to devise solutions, it just acts as a new tool to dig deeper into the infinite dimensions of the nearly known.
from visual import * # In Linux right button mouse drag spins double button mouse drag zooms d = 25 r = 0.1 scene.background = color.white scene.x = scene.y = 0 scene.width = scene.height = 800 scene.range = 1.2*d xaxis = cylinder(pos=vector(0,0,0), axis=vector(d,0,0), radius=r) yaxis = cylinder(pos=vector(0,0,0), axis=vector(0,d,0), radius=r) zaxis = cylinder(pos=vector(0,0,0), axis=vector(0,0,d), radius=r) label (pos=xaxis.pos + xaxis.axis, text='x', box=0) label (pos=yaxis.pos + yaxis.axis, text='y', box=0) label (pos=zaxis.pos + zaxis.axis, text='z', box=0) #look through Earth (@ origin) towards constellation scene.forward = vector (0, 0.8, -0.5) # Plot the stars as spheres -- Earth @ (0.0, 0.0, 0.0), (x, y, z) = light yrs from Earth, radius = arbitrary size Dubhe=sphere(pos=vector(12.2, 92.5, -48.2), radius=1, color=color.yellow) Merak=sphere(pos=vector(11.2,66.6, -42.8), radius=1, color=color.yellow) Phecda=sphere(pos=vector(1.5, 72.5, -53.3), radius=1, color=color.yellow) Megrez=sphere(pos=vector(-2.3,54.5, -35.3), radius=1, color=color.yellow) Alioth=sphere(pos=vector(-9.1,58.0, -38.1), radius=1, color=color.yellow) Mizar=sphere(pos=vector(-18.0,72.0, -47.2), radius=1, color=color.yellow) Alkaid=sphere(pos=vector(-61.6,159.3, -122.2), radius=1, color=color.yellow) #Paths DubheToMerak=cylinder(pos=Dubhe.pos, axis=Merak.pos-Dubhe.pos, radius=0.2,color=color.blue) MerakToPhecda=cylinder(pos=Merak.pos, axis=Phecda.pos-Merak.pos, radius=0.2,color=color.blue) PhecdaToMegrez=cylinder(pos=Phecda.pos, axis=Megrez.pos-Phecda.pos, radius=0.2,color=color.blue) MegrezToAlioth=cylinder(pos=Megrez.pos, axis=Alioth.pos-Megrez.pos, radius=0.2,color=color.blue) AliothToMizar=cylinder(pos=Alioth.pos, axis=Mizar.pos-Alioth.pos, radius=0.2,color=color.blue) MizarToAlkaid=cylinder(pos=Mizar.pos, axis=Alkaid.pos-Mizar.pos, radius=0.2,color=color.blue)