I'm make a big move into Python data processing. Here are some technical notes to myself.
Notes of Parallel Programming with numpy and scipy
http://wiki.scipy.org/ParallelProgramming
My question of Numpy matrix multiplication with custom dot product got answered quickly
http://stackoverflow.com/questions/19278313/numpy-matrix-multiplication-with-custom-dot-product/19278448
How to hash numpy array
Answer: use hash(a.tostring())
http://stackoverflow.com/questions/16589791/most-efficient-property-to-hash-for-numpy-array
I am typing out the Anaconda scipy stack distribution. The download is gigantic 300M file. Then I create a Python 3, environment, which download another few hundred meg of Python 3 packages. I should follow the note on Using Python 3 as my default Python instead.
2013.10.09 comments -