Comments for Ivan Idris Blog

Comment on Steady State Vector of Markov Chains by admin

admin — Tue, 08 May 2012 18:11:27 +0000

Travis,

You are probably correct. Did you use this code? What did you change?

Thanks,

Ivan

Comment on Steady State Vector of Markov Chains by Travis Hoppe

Travis Hoppe — Tue, 08 May 2012 17:48:36 +0000

The correct steady state distribution is:

[ .439877551254698, 0.011573143688891, .548547409068765]

Which you can verify by taking A^k for some large k. Your steady state vector doesn’t even make sense [0.57, 0.57, 0.57]. How can the sum of the probability be greater then unity? In addition, to get a steady-state vector with equal probabilities you’ll need a strong symmetry in your underlying markov matrix (which is also not the case here).

Comment on Stock dynamics and entropy by Andrew Willis

Andrew Willis — Wed, 04 Apr 2012 18:36:19 +0000

How does the normal entropy differ from informational entropy? How is it calculated in this case?

Comment on Christmas NumPy Book Giveaway by Don Davis

Don Davis — Mon, 30 Jan 2012 13:43:33 +0000

I’m interested in trying out NumPy instead of R. I see that the book introduces correlations and histograms. Those would be good starting points. The Markov chain tutorial looked interesting as well.

Comment on Christmas NumPy Book Giveaway by dirk dierickx

dirk dierickx — Fri, 27 Jan 2012 18:20:48 +0000

Hoping numpy will be able to help me at my job by processing, predicting and graphing system/application resources usage for service level management.

Comment on Christmas NumPy Book Giveaway by Amit

Amit — Fri, 27 Jan 2012 16:07:14 +0000

I am a technical writer by choice. I win the book. I read it. And if it impresses me, I shall write about it. And I get to learn NumPy. Winnings everywhere!

Comment on Christmas NumPy Book Giveaway by Juan Manuel Bautista Hoepfner

Juan Manuel Bautista Hoepfner — Tue, 24 Jan 2012 13:56:41 +0000

I am learning Python to make my work more efficient. And from what I have gathered, NumPy makes working with numbers more efficient. So it is all a win-win situation.

Comment on Christmas NumPy Book Giveaway by Lucian Sasu

Lucian Sasu — Tue, 10 Jan 2012 16:06:58 +0000

NumPy offers a natural approach in writing linear algebra code in Python. Created as a free alternative to the expensive Matlab package, it allows for rapid prototyping of a specific class of applications. It is integrated in some specialized packages like Theano (quoting from Theano’s homepage: “tight integration with numpy – Use numpy.ndarray in Theano-compiled functions”) or SciPy. From http://numpy.scipy.org/ it is claimed that it also integrates with legacy code written in C++ or Fortran (yep, old, but a very efficient compiler…) There are plenty of tutorials showing how to use numpy for signal processing and in other challenging areas.

Compared to other numerical targeting packages/languages (like R or Octave/Scilab), the Python + Numpy mix seems to worth investigation: one can use her OOP skills offered by Python in conjunction with a linear algebra package. Developing neural networks or prototyping algorithms in computer vision are supported by Numpy. As Python seems to be the language of choice for developing or prototyping Machine Learning systems (see for eaxmple the answers from http://www.quora.com/What-are-the-best-open-source-machine-learning-libraries-written-in-Python), Numpy is certainly a “must” for anyone interested in developing such softwares.

Comment on Christmas NumPy Book Giveaway by Mykyta Khliestov

Mykyta Khliestov — Fri, 30 Dec 2011 00:07:49 +0000

I’d like to use to improve my understanding of quantum mechanics with simulations.

Comment on Christmas NumPy Book Giveaway by Thomas Ballinger

Thomas Ballinger — Wed, 28 Dec 2011 18:50:47 +0000

I’ve recently been reminded that learning things from books is fun – the structure introduced by an author is helpgul to me.