[Week 02] — Solving Bugs!


Solving bugs for eternity!

To those who have been my blog, it would have been a bit surprising for them for getting this blog-post a bit late. This has been a result of hectic work schedule and a little bit of procastination :P

This blog-post sums up my work for the second week of Google Summer of Code 2019.

Telecon with my mentor

As I might have mentioned(or maybe not), but my other mentor, @DanRyanIrish, was away for a while now, and after discussing with him, we managed to set up a time-slot for a telecon, for getting him updated with my work.

The main points that we discussed were, getting this #PR176 merged. We also discussed, the progress of the main PR for the project, and how much did I make progress. I went through my PR, and discussed a few points about how I refactored the existing methods, and briefly explained my understanding. Code reviews are necessary, and after discussing with Dan we decided to change the logic of some part of the code.

As I had completed most of the tasks for this week, we agreed to start the closed PR. This PR helped in writing a FITS file from a NDCube object. I had started contributing to NDCube with this PR, most of the work had been completed. I had to rebase and write the tests and docs.

Completing the tasks

Throughout this week, I had another discussion with my mentor Dan, who helped me in understanding the code of plotting. The #PR176 was failing some tests, which were part of the plotting of >1D NDCube objects. I had earlier skimmed through the code of plotting but still found understanding the code of plotting quite complex.

Dan helped me in understanding the flow of code in plotting, and it helped me in fast-tracking my work with plotting. Refactoring the tests were much easier, though repetitive. Though the real happiness lies, when all the tests pass, and there is a green color message, signaling that all of your tests have passed.

Thoughts for the next week

With most of the tasks completed for refactoring NDCube with APE14 methods, it was now clear for me to take on the second stage of my project — Slicing of NDCube. Slicing by far was presumed by my mentor @Cadair as the most complex part of the project. I was apprised by him about how the slicing can take most of the time of the project, so I was rather aware of that. I am planning to start playing with the slicing, and how it used to work for NDCube earlier, and how the behavior changes with APE14 changes. This should take a fair amount of time, so I would hope that this gets completed within a reasonable amount of time.

Link to my previous post

  • [Week 01] — Getting Stuff Done! — Link

    Until the next blog-post, ciao-adios!


comments powered by Disqus