My latest work involves upgrading the approximations implemented in a lot of the evalf() methods, along with adding some new approximations. Problems with precision caused the older evalf() methods to evaluate incorrectly. Although the error was little, to someone actually asking for high precision results would probably consider it quite significant. There was a lot of work to be done, but I have most of it working now, albeit some of the new series approximations I have implemented converge slowly in large domains, they are at least there as a foundation. When working on this I came across a lot of small bugs that were floating around and fixed them also.

After finishing this I was able to use arc cosine approximations to ensure that Polygon.angles was returning correct results (or at least for the test cases I wrote it works!). After I get a few more test cases written I'll be committing my latest Geometry module work and then consider working on some more core stuff.

No comments: