class Core
package geometrize
The core functions in Geometrize Haxe.
Static methods
staticbestHillClimbState (shapes:Array<ShapeType>, alpha:Int, n:Int, age:Int, target:Bitmap, current:Bitmap, buffer:Bitmap, lastScore:Float):State
Gets the best state using a hill climbing algorithm.
Parameters:
shapes | The types of shape to use. |
|---|---|
alpha | The opacity of the shape. |
n | The number of random states to generate. |
age | The number of hillclimbing steps. |
target | The target bitmap. |
current | The current bitmap. |
buffer | The buffer bitmap. |
lastScore | The last score recorded by the model. |
Returns:
The best state acquired from hill climbing i.e. the one with the lowest energy.
staticbestRandomState (shapes:Array<ShapeType>, alpha:Int, n:Int, target:Bitmap, current:Bitmap, buffer:Bitmap, lastScore:Float):State
Gets the best state using a random algorithm.
Parameters:
shapes | The types of shape to use. |
|---|---|
alpha | The opacity of the shape. |
n | The number of states to try. |
target | The target bitmap. |
current | The current bitmap. |
buffer | The buffer bitmap. |
lastScore | The last score recorded by the model. |
Returns:
The best random state i.e. the one with the lowest energy.
staticcomputeColor (target:Bitmap, current:Bitmap, lines:Array<Scanline>, alpha:Int):Rgba
Calculates the color of the scanlines.
Parameters:
target | The target image. |
|---|---|
current | The current image. |
lines | The scanlines. |
alpha | The alpha of the scanline. |
Returns:
The color of the scanlines.
staticdifferenceFull (first:Bitmap, second:Bitmap):Float
Calculates the root-mean-square error between two bitmaps.
Parameters:
first | The first bitmap. |
|---|---|
second | The second bitmap. |
Returns:
The difference/error measure between the two bitmaps.
staticdifferencePartial (target:Bitmap, before:Bitmap, after:Bitmap, score:Float, lines:Array<Scanline>):Float
Calculates the root-mean-square error between the parts of the two bitmaps within the scanline mask. This is for optimization purposes, it lets us calculate new error values only for parts of the image we know have changed.
Parameters:
target | The target bitmap. |
|---|---|
before | The bitmap before the change. |
after | The bitmap after the change. |
score | The score. |
lines | The scanlines. |
Returns:
The difference/error between the two bitmaps, masked by the scanlines.
staticenergy (shape:Shape, alpha:Int, target:Bitmap, current:Bitmap, buffer:Bitmap, score:Float):Float
Calculates a measure of the improvement adding the shape provides - lower energy is better.
Parameters:
shape | The shape to check. |
|---|---|
alpha | The alpha of the shape. |
target | The target bitmap. |
current | The current bitmap. |
buffer | The buffer bitmap. |
score | The score. |
Returns:
The energy measure.
statichillClimb (state:State, maxAge:Int, lastScore:Float):State
Hill climbing optimization algorithm, attempts to minimize energy (the error/difference).
Parameters:
state | The state to optimize. |
|---|---|
maxAge | The maximum age. |
lastScore | The last score recorded by the model. |
Returns:
The best state found from hillclimbing.