GSoC Week 13

12 Aug 2016

This week I spent a lot of time working on FeatherstonesMethod and its component parts. I started off by moving a bunch of spatial vector functions from another PR I have to the featherstone PR and used some of those functions to calculate the spatial inertia of Body objects. The next thing I worked on was completely rewriting the internals of the joint code. The joints now consist of 4 reference frames and points (one set at each of the bodys mass centers and one set per body at the joint location).

After this I ran some basic code that used these new features and kept making changes until the code was able to run without producing errors. I used this same method of work with FeatherstonesMethod and now it too is able to run without producing errors. Now that the code runs it was time to make sure that the output is correct which is a lot more involved than the previous step of work. To begin I solved for the spatial inertia by hand and used this calculation to create test code for Body.spatial_inertia. As expected the code initially was completely incorrect but it now passes the test. I have since been working on the tests for the joint code. Since this code is completely new to the sympy repository it takes a lot more planning than the body test did. Also I need to solve the kinematics by hand for the joints so that I have a base for the test code. This is where I am currently located in the process.

Also this week I addressed review comments on SymbolicSystem and have moved that PR closer to being able to merge. One of the current hang ups is trying to force Sphinx to autodocument the __init__ method. I think the best solution currently is to move the relevant code back to the main docstring for the class and not worry about trying to document the __init__ method.

While working on rewriting the joint code I came across a bug in and have created a docstring with a fix to this along with a test to make sure the fix works.

Lastly I reviewed a PR that adds a docstring to a method that did not yet have a docstring. The PR had some information in it that was incorrect and after some research I was able to make some suggestions for its implementation.

Future Directions

Next week is the last full week of GSoC and my main priority is getting the final evaluation information correctly finished so that the work can be processed correctly. My next goal is to make sure SymbolicSystem gets merged into SymPy. This is not entirely in my hands, however, as I will be having to wait for feedback and so while waiting I will be pulling off different parts of FeatherstonesMethod for separate PR’s at the recomendation of my advisor. These separate PR’s I hope to possibly include in my final evaluation.

PR’s and Issues

  • (Open) [WIP] Added to physics/mechanics PR #11431
  • (Open) [WIP] FeatherstonesMethod PR #11415
  • (Open) Added docstring to jordan_cell method PR #10356


Want to leave a comment? Visit this post's issue page on GitHub (you'll need a GitHub account).