I don't think I follow on the first part. Edinburgh says that the complexity of the two traversals are dramatically different. repeat for each is somewhere between nlogn and n, and repeat with is n^2. At least for the case of your squares of integers, I would expect that there is a crossover where it's going to be faster to build the list, first. I don't know if that is at 100, 1000, or some bigger number, but n vs. n^2 is a very big difference.
On Mon, Feb 16, 2015 at 2:51 PM, Ben Rubinstein <benr...@cogapp.com> wrote: > On 15/02/2015 02:28, Mike Kerner wrote: > >> I just read the dictionary entry (again), and I would say that it is not >> at >> all clear that there would appear to be an ENORMOUS difference. For >> starters, you have to read waaaaaaaay down to find the mention, it isn't >> really called out with a NOTE or anything else to draw one's attention to >> it, and it is definitely understated. Even mentioning "order of >> magnitude" >> would be better (although it would appear to be an understatement). I >> literally had no idea until I ran into this, by accident, and was >> exchanging notes with Peter. The difference is staggering, and it really >> should be made much more obvious. >> > > I think that we need to be clear that it's not true that "repeat with" is > massively faster than "repeat for". They're not comparable, they're doing > different things. > > What _is_ true is that if you want to iterate over the chunks of a > variable, the most efficient way to do that is using the "repeat for each" > construction. But there's nothing inherently faster in the loop form, it's > what you do with it that counts. If you wanted to print the squares of the > first 100 integers, "repeat with n = 1 to 100" is the way to go - don't > make a list of those 100 integers so that you can use "repeat for each" on > it. > > I'm not saying the documentation is perfect, far from it. But what it > says on this topic is accurate, i.e. "*if* you want to perform an action on > each chunk in a container. This form is much faster than ... with". > > On 14/02/2015 22:00, Mike Kerner wrote: > >> You would think that Edinburgh would think about tweaking an algorithm, >> since REPEAT WITH seems to be a special case of REPEAT FOR, and you can >> generate the REPEAT WITH behavior by wrapping the REPEAT FOR... >> > > That's exactly what (Scott Raney) did, back in the mists of time. He > wrote extra code to make iterating over the chunks of a variable > exceptionally fast; and added syntax for the programmer to indicate that > they wanted to use this algorithm (which has other disadvantages). > > REPEAT WITH is certainly not a special case of REPEAT FOR EACH (note > 'repeat for' is actually something else); it's closer to the other way > round, that is REPEAT FOR EACH could be regarded as a specialisation for a > common case otherwise handled by REPEAT WITH. > > Ben > > > _______________________________________________ > use-livecode mailing list > use-livecode@lists.runrev.com > Please visit this url to subscribe, unsubscribe and manage your > subscription preferences: > http://lists.runrev.com/mailman/listinfo/use-livecode > -- On the first day, God created the heavens and the Earth On the second day, God created the oceans. On the third day, God put the animals on hold for a few hours, and did a little diving. And God said, "This is good." _______________________________________________ use-livecode mailing list use-livecode@lists.runrev.com Please visit this url to subscribe, unsubscribe and manage your subscription preferences: http://lists.runrev.com/mailman/listinfo/use-livecode
