Timing puzzle - could Sow and Reap help?
- To: mathgroup at smc.vnet.net
- Subject: [mg130466] Timing puzzle - could Sow and Reap help?
- From: carlos%colorado.edu at gtempaccount.com
- Date: Sat, 13 Apr 2013 02:02:51 -0400 (EDT)
- Delivered-to: l-mathgroup@mail-archive0.wolfram.com
- Delivered-to: l-mathgroup@wolfram.com
- Delivered-to: mathgroup-newout@smc.vnet.net
- Delivered-to: mathgroup-newsend@smc.vnet.net
Thanks to all who reply to my post. I understand now why there is no Array head. Array seems to be just a symbol building function. Lists are stored as arrays, a la SLIP (a now dead language created by Joseph Weizembaum at MIT, which I used briefly in the early 1970s). Separating Lists and Arrays would make a significant improvement in efficiency in getting the best performance from each data structure. But that would take a significant kernel rewrite. Going back to the test timer, I have never used Sow and Reap, as suggested. Could somebody illustrate how to built the test plot list with those? Then I can time it against the first 4 methods (as noted, any scheme using Table or Map is useless in practice since n is not known in advance). Thanks.