![]() THE CYTOSCAPE CONSORTIUM, SAN DIEGO, CA : 3,312.539, PUB. Each concern can be addressed in a separate issue, cross-referenced with this issue. THE COCA - COLA COMPANY, ATLANTA, GA : 3,313.553. Latest version: 2.5.1, last published: a year ago. More details are provided in #2473.Ä«ecause there are three separate concerns raised by your demo, I'm going to leave this issue closed. 0.8.4 cytoscape-cola 2.3.0. The Cola.js physics simulation layout for Cytoscape.js.The handling of zero-area elements - especially for layout extensions that are third-party or that consume third-party libraries for the layout algorithm - are not as robust as they should be. fitting in a sensible way: Zero-area elements are excluded from the composite bounding box. (3) The library already handles zero-area (e.g. Usually the intention there is to run a layout once on the entire graph so that showing a previously-hidden element requires no new layout: The previously-hidden element is already in place. Now one click is all you need to arrange your Cytoscape networks in a clear and concise way with the yFiles layout algorithms and edge routing algorithms at. Having said that, there are valid use-cases for running a layout on hidden elements. Click any example below to run it instantly cytoscape-layout-tests. However, revising the library will not solve the root issue in your demo: You're getting bad layout results, since you're running a layout on hidden elements. Use this online cytoscape-fcose playground to view and fork cytoscape-fcose example apps and templates on CodeSandbox. That commit does not use ele.takesUpSpace() in the bounds functions, which would be necessary to be consistent with the intention of that commit. cytoscape-cola examples - CodeSandbox Cytoscape Cola Examples and Templates Use this online cytoscape-cola playground to view and fork cytoscape-cola example apps and templates on CodeSandbox. (2) There is indeed a problem with the library itself, which is demonstrated by your demo, introduced in commit 47360ca. Even if there were no exceptions in your demo (as in v3.4.2), the use of the API within that demo isn't right - not yielding the intended behaviour. ![]() It should also be clearer that you can run a layout on a subgraph, and common use-cases should be given for when that would be appropriate (e.g. It should be clearer what the semantics are for different ways of hiding elements. (1) The main issue is a failure of the documentation. I've reviewed this further and I've come to three conclusions: The restrictions on fitting are similar to those of layout for this discussion.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |