Community Bonding - A head start
So, here I am with another set of experiences to share with you, i.e., the Community Bonding of GSoC. After discussing with my fellow developers, and my mentors, we reframed our timeline. I will be working on very interesting topics in statistics, including, joint distributions, Markov chains, random matrices, assumptions of dependence and simplifying the results of the stats module.
For a head start, I started working on adding multivariate distributions and the good news is that I have completed majority of the work, apart from some technical constraints which will be solved by another project very soon. If you are interested in seeing the code, take a look at the PRs #16576, #16808, #16825. Infact, apart from the stats module, I also had an opportunity to work on other modules of SymPy, like, I improved the API of Sum
in sympy.concrete
, allowing Range
to be passed as limits and that too while working on discrete joint distributions. You can check the PR #16810 for the details. Thanks to Christopher Smith for helping me with this. Still there are issues like, #16833 which require a lot of discussion before concrete implementation, will update you on this if some progress takes place.
Apart from coding, we had discussions on design of Markov chain and random matrices. I also made two prototypes at PR #16852, and #16866 . We are currently having more to think on this part, like API, class structure, etc.
Overall, the bonding period was very a great learning experience about how various modules affect each other and bring the best out of SymPy. See you soon in the next blog. :)