Skip to main content


Showing posts from January, 2015

Graphene : Elasticity meets electronics on atom-thin membranes

The speaker at the physics colloquium this week was Prof. Manu Jaiswal from the Dept. of Physics, IITM and his talk was about graphene, exciting research on the elasticity of graphene coatings/layers and on the recent developments in the field.

The talk started off with Zidane's penalty kick of the football world cup and at the end, he pointed out how the shape of the goal net has changed. He mentioned that physicist and mechanical engineers aren't the only people working on material elasticity, so are footballers, so to speak. The goal net design changed in 2002 from being a square net to a hexagonal net. It was shown, through simulations, how stress propagates along the two structures and how hexagonal structures behave like a 2D isotropic body where as square nets have local minima along the grid outlines.

Moving on the talk, he mentioned how he himself got involved in understanding the elasticity of graphene. As is the case with many-a-stories in science, results that cou…

Variable stars from Stripe 82.

I finally uploaded the ipython notebooks related to variable stars, the thing that I was working on over the weekend. And I also started converting a solution to the ant colony optimization problem that my and friends had worked on a couple of years to python from matlab. We had worked on it as part of a competition, that we ended up actually winning. I am a still working on it but you can access it here. I just figured out that I should look through the file containing object name and period to search for objects with small periods and then plot those light curves, to see variability.

the weekend that was

The weekend was pretty much spent writing more code in fortran, trying to get accustomed to the syntax and thinking in that language. That's another interesting thing that I noticed now that I'm writing the same code in multiple languages. Not only is the syntax and the code different, the thought process in itself which leads to the code is different. I guess this is one of the reason people argue as to which programming language is better and which is worse, depending on the thought process it stimulates in a programmer. I'm still getting used to the thinking in fortran. And it's been an interesting road too. I couldn't find useful resources online to help me start writing code in fortran, I guess I didn't search for long enough or hard enough. Either way, I ended up looking up code passed on to me by my professors, which I had blindly used as part of my data analysis routine, for syntax and usage. It was a bit like learning a language by reading the sign bo…

Software baggage

Most of my afternoon was spent downloading GalSim, which failed twice for reasons that are beyond me! And I have been spending the rest of the time installing it's dependencies, which as you can see here, are mind boggling. Honestly, the sheer number of dependencies, python and otherwise, is making me think twice about what I'm getting myself into but well, I've started digging, might as well see it through.

While I was waiting for the download and for the installation, I figured out how to write a working fortran file! Just one that measures the area of a circle given the radius as input, which can then be saved to a file. I don't understand why every new line starts with a tab character. I now need to get used to control statements in fortran and writing subroutines to be called in the main function. I'm slowly, but surely, updating my reference files for c and fortran syntax.

Woah, it doesn't look like I got much work done today, I'll have to pick up my…

Programming languages is quirky

I've been writing more code lately, and I see myself writing a lot more over the next year. I need to get good at writing C code, keep getting better at python and add fortran to the list of languages that I'm competent in working in, along side the more scientifically oriented matlab and mathematica. I wasn't really into programming in the first few years of my college life although I was interested in algorithms. They were logical and intuitive and thought provoking, which isn't usually the case with programming languages. I guess they're like spoken languages and the same way the phonetics of a language and the underlying rules need to be respected, if you want the other person to understand what you're trying to convey, a program has the various semi-colons, colons, braces and everything so the computer, actually the compiler, understand what you want it to do. I brought up the spoken language analogy because i'm trying to think if there is a programmi…

A new game plan

Based on what I mentioned a couple of days back, I'm now trying to put together the basics of programming in C and Fortran, for a course I'm doing this semester. Pretty much most of my evening was spent on those snippets on how to code, compile and debug. Earlier I attended a popular science talk by Prof. Sriram on how supernovae were used to discover an accelerating expansion of the universe. He introduced the concept of the cosmic distance ladder and he talked about how supernova, of a specific kind, can act as standard candles, that can be used to measure distances to far away galaxies.
And I had an idea for another interesting side project. Well, I already had the idea like an year back but never took it up seriously. I wanted to understand how a random distribution of particles end up on a plane, look at the milky way galaxy. There should be a way to analytically solve and understand where such a plane might lie. It would also be an interesting visualization to show an n…

Jupiter in JPEG

Over the weekend, me and a couple of hyper active juniors tried taking pictures of various celestial objects, over the course of a couple of hours. We tried Jupiter, Pleiades, Orion and the comet Lovejoy. It was fulfilling to have atleast gotten some pictures of jupiter, although they were (slightly) out-of-focus. We used a DSLR and we didn't use the ability of our mount to track the object for a longer exposure, something that we definitely need to try the next time!

I just got my hands on the data in the evening and I wanted to try something different, different than the usual image processing done on Jupiter and the like. I wanted to see if i can get any useful information out of the picture, like the ratio of distances between the moons of jupiter or the ratio of radius of jupiter to distances of various moons. But for that I need to identify the moons and their *centers*. Here are the results of my initial attempt at modeling the out-of-focus moons and I just realized that a…

Everyday lives of everyday people

If you will humor me for a moment, when I was in my second year of college, I was part of the Extra Mural Lecture team. Over the academic year, we invited people from different facets of life to give talks and inspire students to look outside the confines of our campus. One of the speaker we had invited that year was Mr. P. Sainath and for various reasons, I ended up not going to his talk. I then got an earful, from a lot of people, praising the talk and balking at my absence. So I wasn't going to miss him a second time when he came today, to deliver the pre-conference talk.
Right off the bat, he talked about the mass migration happening in rural India towards urban and metros. He talked about the number of farmer suicides over the last 20 years and he talked about the amount by which the number of full time farmers had fallen while, at the same time, the number of part time agricultural laborers increased. He talked about the shrinking communities in rural India and how this is …

Astronomical data - object coordinates

As part of the workshop on astronomical data analysis that I helped organize during Shaastra '15, we introduced participants to celestial coordinates and how to query for object coordinates using sql query on the SDSS sql query page. Here's a handy ipython notebook that runs through the sql query code, show you how the data set queried for looks like and how to load the data and plot it using python. It also gives two different perspectives of the same data set i.e one plot made using the equatorial coordinates and the other made using galactic coordinates.

PS - if you work on the terminal regularly, I suggest you look up what the reverse-i-search is and how it can make your life on the terminal easy!

An (practical) introduction to programming (in python)

I have been learning python for almost two years now, mostly for astronomy oriented research projects but over the course of the last semester, for a more diverse set of academic questions, as part of a course. I have come across quite a few nuggets that have helped make my life easier while working on python. Looking at the big picture, I have always liked writing about the things that I learn and I feel that only after writing about what I have learnt, in a meaningful and relatable fashion, do I really understand the topic. Quite a few of you might've come across how asking questions out loud, even if it is to a wall, helps understanding the question and see it from a different perspective. Trying to teach someone else what you have learnt will need you to arrange and order things in your head, so that it makes sense to the other person and I don't have to tell you how this will help you understand the topic better as well. Coming back, I have been trying to put together th…

Physics colloquium on CERN's LHC

The semester has started and the first colloquium in the Dept. of Physics here at IIT Madras happened today. The speaker for today was Dr. Kajari Mazumdar, a faculty at TIFR, Bombay. She talked on the topic 'LHC : The unbelievable pursuit of the unimaginable' and over the course of her talk, she described how the unimaginable physics (to most of the public) is being brought to life using unbelievable machines and brain power.

Personally speaking, I have come to admire speakers who put numbers in their popular talks. Early on in my under graduation here at IITM, during a popular session on astronomy, I was asked to stop speaking in qualitative terms and be  quantitative, especially given my interest in pursuing physics after graduation. I have come to understand how beautiful stories can be woven around numbers and have come to admire professors who give a better perspective of their work and the research area in general by quantifying it in a way the layman can understand.

To be…


As you can obviously see, things have been running a bit slow lately. Shaastra and Saarang, the annual technical and cultural festivals at IIT Madras were organized over the last week. I tried participating in the Paper & Poster presentation event and in the Industry Defined Problem event but I didn't qualify in neither of the events and the EDM night was the only thing in Saarang that I wanted to attend. The rest of the time was spent either having meals or waiting for the next meal. The institute has officially reopened today and I need to get my chickens in order, decide whether or not I want to add extra courses and so on. Midst all of this, I'm wondering when I'll get to work on the ideas that are bouncing around in my head, afraid that what they say is true, that original thoughts some how vanish after the 30's. Well, I'll just have to make the most of the time I have I guess. Managing time is something that I've had trouble with over the years and I …

Working with fits files

As I had mentioned, much much earlier, almost all data files, be it photometric or spectroscopic data, are stored in .fits format. The pyfits library or the astropy library can be used to read and write such files in python and the cfitsio library can be used to work on them in C. As part of a small workshop on astronomical data analysis, I showed students how astropy can be used to extract data from .fits files, containing information about the spectrum of an astronomical object. You can find the relevant fits files, ipython notebook and a html version of the ipynb on my github repository. I shall include data regarding image fits files tomorrow.

Books! Books! Books!

I don't like making new year's resolutions. I don't understand what is special about that particular day that people believe will suddenly help them keep the resolutions they make. Sorry to break your bubble people but if you want to make resolutions and change your life, start now instead of waiting for a particular day.

Anyway, enough philosophy. Around the time when I wanted to write daily, I wanted to start reading again. I haven't read books in a long time. In a very long time. For whatever reason, it appears to people as if I read a lot. Well, I do infact read a lot but most of them are articles online, some of them tens of thousands of words long but none of them actual books. When I was a kid and till my tenth standard, I used to read books about machines and how they work, basic science and year books briefing what all happened over the course of the year. After I got into IITM, the only book that I held in my hands for longer than a couple of hours is 'E…

Augmented reality devices in today's workplace

Industry defined problems is a regular event at Shaastra, the annual technological event at IIT Madras. Industries sponsor the event and students compete to give interesting solutions to said problems. This year, EATON sponsored the event, defining 3 problems statements for those interested to tackle. I was interested in trying my hand at one of the problem statements, on how augmented reality devices or specifically the Google Glass can help multi-national, multi-product, technology oriented industries to increase productivity.

In the following presentation, I talk about the ways in which certain industries can flourish with the use of augmented reality devices.

To elaborate on the four important topics mentioned in the presentation, the first of which is classroom education, a large portion of students in schools are put off by science and math because they cannot visualize the subjects and therefore, have a poor understanding on the subjects. Topics such as three dimensional vector…