Comments on: MapReduce in 10 or so lines of Python http://www.chencer.com/dave/blog/2008/07/09/mapreduce-in-10-or-so-lines-of-python/ David Spencer's personal weblog Wed, 16 Feb 2011 06:42:44 +0000 hourly 1 http://wordpress.org/?v=3.0.5 By: dave http://www.chencer.com/dave/blog/2008/07/09/mapreduce-in-10-or-so-lines-of-python/comment-page-1/#comment-126494 dave Mon, 27 Apr 2009 05:39:25 +0000 http://www.tropo.com/dave/blog/2008/07/09/mapreduce-in-10-or-so-lines-of-python/#comment-126494 Yes, nice, thank you, this is much more logical, concise, crisp, and in the spirit of things. Yes, nice, thank you, this is much more logical, concise, crisp, and in the spirit of things.

]]> By: Amit Patel http://www.chencer.com/dave/blog/2008/07/09/mapreduce-in-10-or-so-lines-of-python/comment-page-1/#comment-126493 Amit Patel Sun, 26 Apr 2009 17:12:38 +0000 http://www.tropo.com/dave/blog/2008/07/09/mapreduce-in-10-or-so-lines-of-python/#comment-126493 I think of MapReduce grouping the results as the mapper produces them: def MrSimpler(producer, mapper, reducer, consumer): … sharded = {} … for n, v in producer(): … … for key, value in mapper(n, v): … … … sharded[key].setdefault([]).append(value) … for key, values in sharded: … … consumer(key, reducer(key, values)) I'm not sure what the actual implementation does though. I think of MapReduce grouping the results as the mapper produces them:

def MrSimpler(producer, mapper, reducer, consumer):
… sharded = {}
… for n, v in producer():
… … for key, value in mapper(n, v):
… … … sharded[key].setdefault([]).append(value)
… for key, values in sharded:
… … consumer(key, reducer(key, values))

I’m not sure what the actual implementation does though.

]]>