This is my first post and hence I begin by congratulating myself on getting selected for the Google Summer of Code. I will begin by describing the organization I got selected for : PGMPY. So the pgm in pgmpy stands for probabilistic graphical models and the py stands for python. You would have guessed it by now, but still, it is a organization under the PSF (Python Software Foundation) and works on probabilistic graphical models. Here is the link to the organization https://github.com/pgmpy/pgmpy . Sadly it doesn't still have a "web-page" of its own but still the github page will do. It is relatively a very new organization and they have very little code right now (as compared to many other organizations. A lot of previous GSOCers said that it took them quite a while to get familiar with just the part of the code which was relevant to them).
Anyway, let's get to what I will be doing. My task will be to basically implement the Markov Networks. After implementing the basic architecture, I will be implementing a few algorithms. Take a look here for the current version of the timeline (We decided to make some changes to the initial timeline which I had proposed. A few changes here and there. Though it would hardly be interesting for you ) https://docs.google.com/document/d/1WRuiEUgiuwWMuxezTy-YXqSiVILP2Q1lQ50zU6kib6g/edit?usp=sharing .
Now, the big question is what I have been doing since 19th of April to today . (Which is approximately a month). We were not supposed to start coding yet as it is still the community-bonding period and we are supposed to familiarize ourselves with the code, become friendly with the mentors, get to know people in the org, and do any preparatory work, as necessary to begin the coding work. So I haven't yet started coding up stuff. However I have been doing some cool stuff.
To being with, I had my end-sems exam till 30th of April and so, I was quite inactive till then. We had an IRC chat on 28th of April and then , we just had a brief chat about things and set a few plans for community bonding period. We also chatted a bit about IDE and cython and testing-techniques among other things. After the exams, I lazed away for a while and after about the 3rd of May started reading up the algorithms which I would be implementing. I did a course on Advanced Machine Learning this semester , however, I thought that it would be a good idea to go through the algorithm from the book (the standard book on Probabilistic Graphical Models ofcourse, the one and only, Probabilistic Graphical Models, by Koller and Friedman).
We had the next IRC chat on 10th May (oh , we had decided in the previous IRC chat that we would be having these chats every Saturday starting from 10th). Then it was more of a report of progress and clearing doubts, if any. We discussed plans about the subsequent weeks. So, as per the plan, I went through the software preparations part of my project, and familiarized myself with Cython, PyCharm IDE, configured VIM to work well with python (that was quite a struggle, but then, anything for VIM :D ). I also took a look at the exisiting code so that I could use parts of it for my implementation. (I also guess that I found a bug , for which I sent a pull request).
And then we land up at today. We had another IRC chat yesterday and my task for the next week would be setting up the Markov Networks class and writing the basic class implementation, testing it up and so on. (Look at the timeline) But the good news is that, coding starts now. :)
Anyway, let's get to what I will be doing. My task will be to basically implement the Markov Networks. After implementing the basic architecture, I will be implementing a few algorithms. Take a look here for the current version of the timeline (We decided to make some changes to the initial timeline which I had proposed. A few changes here and there. Though it would hardly be interesting for you ) https://docs.google.com/document/d/1WRuiEUgiuwWMuxezTy-YXqSiVILP2Q1lQ50zU6kib6g/edit?usp=sharing .
Now, the big question is what I have been doing since 19th of April to today . (Which is approximately a month). We were not supposed to start coding yet as it is still the community-bonding period and we are supposed to familiarize ourselves with the code, become friendly with the mentors, get to know people in the org, and do any preparatory work, as necessary to begin the coding work. So I haven't yet started coding up stuff. However I have been doing some cool stuff.
To being with, I had my end-sems exam till 30th of April and so, I was quite inactive till then. We had an IRC chat on 28th of April and then , we just had a brief chat about things and set a few plans for community bonding period. We also chatted a bit about IDE and cython and testing-techniques among other things. After the exams, I lazed away for a while and after about the 3rd of May started reading up the algorithms which I would be implementing. I did a course on Advanced Machine Learning this semester , however, I thought that it would be a good idea to go through the algorithm from the book (the standard book on Probabilistic Graphical Models ofcourse, the one and only, Probabilistic Graphical Models, by Koller and Friedman).
We had the next IRC chat on 10th May (oh , we had decided in the previous IRC chat that we would be having these chats every Saturday starting from 10th). Then it was more of a report of progress and clearing doubts, if any. We discussed plans about the subsequent weeks. So, as per the plan, I went through the software preparations part of my project, and familiarized myself with Cython, PyCharm IDE, configured VIM to work well with python (that was quite a struggle, but then, anything for VIM :D ). I also took a look at the exisiting code so that I could use parts of it for my implementation. (I also guess that I found a bug , for which I sent a pull request).
And then we land up at today. We had another IRC chat yesterday and my task for the next week would be setting up the Markov Networks class and writing the basic class implementation, testing it up and so on. (Look at the timeline) But the good news is that, coding starts now. :)
No comments:
Post a Comment