Hello friends,
As the midterm evaluation is approaching, I pushed the pedal to the metal and 
my blog and github profile have seen a lot of activity recently.

I would like to link to everything from one place, and that place will be this 
e-mail. So this is what happened:

-- I wrote a couple of magic functions for Fabian's memory_profiler:
  - Phase 1: 
http://blog.vene.ro/2012/06/30/quick-memory-usage-benchmarking-in-ipython/
    This is where I started working on the magic %memit function, and ran into 
the problem of
    needing to run things in a new, clean process.

  - Phase 2: http://blog.vene.ro/2012/07/02/more-on-memory-benchmarking/
    This is where I backported %memit to the older IPython magic style, added a 
%mprun clone
    of %lprun, and packaged everything within my fork of memory_profiler, which 
will be
    merged if it looks rather good to you guys. (Thanks for the vote of 
confidence Fabian!)
    Then for some reason I found that %memit was failing on sklearn functions.

  - Phase 3: http://blog.vene.ro/2012/07/04/on-why-my-memit-fails-on-osx/
    With Olivier's help I pinned this down back to an old bug-friend of ours. 

-- I brought this into vbench and updated the scikit-learn-speed benchmark 
suite to use it.
  - http://blog.vene.ro/2012/07/05/memory-benchmarking-with-vbench/

-- In parallel, I got two speedup pull requests works in progress.
  - Linear models: https://github.com/scikit-learn/scikit-learn/pull/913
  - Decomposition: https://github.com/scikit-learn/scikit-learn/pull/927


Links to my github repositories:

-- My fork of memory_profiler:
   https://github.com/vene/memory_profiler

-- My fork of the scikit-learn-speed project:
   https://github.com/vene/scikit-learn-speed

-- My fork of vbench, the branch supporting memory usage benching:
   https://github.com/vene/vbench/tree/memory


The current short-term priority, in my opinion, is to make use of all this, and 
put together
benchmarks for linear_models, decomposition, and everything else. Then, we can 
figure out
a way to deploy scikit-learn-speed on a remote machine somewhere.

Best,
Vlad



------------------
Vlad N.
http://vene.ro




------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Scikit-learn-general mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/scikit-learn-general

Reply via email to