There are also graph visualizations that have way less features than Cytoscape.js but are faster, so if you don't need any of the features and just want to visualize a simple graph, you may try ngraph, see a demo at. Specifically, you define the x and y coordinates like: let options = I don't know what you mean by "circulate the x,y positions in Java", do you mean "load a preset layout in the JavaScript(!) library Cytoscape.js"? If so, that is explained here. Does the graph at least change infrequently? Then you can cache the layout and only recalculate it if necessary.Is the graph always the same? Then you could precalculate the layout and load it as a preset layout.Also, as far as I know, GPU computing is harder to do in JavaScript.īoth of those problems may be addressed in the future and the developer of Cytoscape.js, Max Franz, seems to be extremely active and supportive, so if JavaScript ever gets better parallelization and GPU computing support, I am positive this will find its way into Cytoscape.js shortly. The problem is that JavaScript is slower in tasks like graph layouting because modern CPUs have more and more cores and JavaScript's parallelism implementation (web workers) have too much overhead for algorithms with many short iteration steps where the result of all threads has to be integrated. Even for medium sizes like your 5000 nodes and edges, 15 seconds sounds normal for Cytoscape.js. terrabytes or more) because it runs in the browser. while Cytoscape apps provide highly performant and relevant network biology functionality, the specialized pro-gramming talent and relatively long development times they require can make them uneconomical for delivering complex and evolving workflows. Cytoscape.js is impossible to be used with actual big data (i.e.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |