motionblur | Generate authentic motion blur kernels ( point spread | Computer Vision library

You are adding the elements to the graph after the layout has been run. That's why all your nodes are piled up on the upper left corner.

You have two choices here:

1. Supply both the layout at the elements on the constructor.
2. Initialize the graph first, then add the elements (e.g. using cy.json()) and then tell cytoscape to run the layout with cy.layout(...).run().

Documentation for cy.layout() is here, and for cy.json() here.

The following snippet uses the first approach, but I have left the second one commented in case you cannot supply the elements on the constructor call (i.e. if you have to fetch them asynchronously).

Here is the solution: fx:controller="application.FXMLDocumentController"

Why not have a different class for the points in the sparkle's trail? Call each point a SparkleTrail, have them be stationary and let them draw themselves with slowly-decreasing opacity, and finally make them vanish (= remove it from the draw() loop) once their opacity reaches 0.

Proposed changes:

• new SparkleTrail class, has two coordinates, an initial opacity of, say, 0.5. Paints itself when you call its draw method. Has a slowlyVanish method that decreases opacity by 0.01 whenever you call it, which you will do each 1/50th of a second (or whatever your framerate is), therefore slowly vanishing after aprox. 1 second.
• create them every 10 frames of moving your Sparkle, adding them with the current sparkle's position to a trail arraylist. This will make the trail brighter when the sparkle moves slow, and space it apart when it moves very fast.
• call draw on all SparkleTrail instances in the trail arraylist whenever you redraw everything. Make sure to call their slowlyVanish method, and when opacity reaches zero, remove them from the trail.

Turns out I needed to initialize in document ready and cy.add([data]) after.

I think you want what's called a "comet" kernel. I'm not sure what kernel is used for the "motion" blur, but I'd guess that it's symmetrical based on the image you provided.

Here is some code to play with that applies the comet kernel in one direction. You'll have to change things around if you want an arbitrary angle. You can see from the output that it's smearing in one direction, since there is a black band on only one side (due to the lack of pixels there).

Without more information it is hard to say what is going on but I would recommend that you add RequestExecutionLevel User to your script to make sure Windows does not apply compatibility shims to the installer.

InstallDirRegKey can trick you if you installed to Program Files during testing on your machine. Make sure that value is not present in the registry when debugging issues like this.

first, you shouldn't just let the primaryStage show() in the for-each loop, because when the primaryStage first show(), the fx thread pause on that instruction, the remaining pictures will not be read and when you close the window, the loop won't continue to go on for remaining, it represents the end of the application.

so it's better to let the primaryStage show() outside the each-loop after all image has been read.

second, you shouldn't use the "Group" container to contains all images, better use VBox/HBox/FlowPane, make sure above image won't covers below's.

here's modified code for reference.