Programming puzzles: processing lists!

[pier4r]

(04-20-2017 09:35 PM)pier4r Wrote:  Challenge #1a userRPL, 50g.

more functions: http://www.hpmuseum.org/forum/thread-8114.html

I noted that the test failed almost immediately and so I started to estimate the probabilities.

If I'm not mistaken, the test will succeed - given a random list - with probability:
1/2 for 2 elements
1/4 for 3 elements
1/8 for 4 elements
1/2^(n-1) for n elements

so it is very likely to fail.
....

So after 2048 tests with a list of size 10 (output equal to 11 when valid).
1st: a lot of memory used, especially to save the results instead of deleting it accidentally.
2nd: the smaller the list, the faster the stack operations.
3rd: I got 2 elevens (list passed completely) and 7 tens (list failed at the last element)
The theory should have been 4 elevens and 8 tens, so I would assume that my code is not failing. To be sure as soon as possible I will code a function that instead of producing randomly a list of two values, will produce with a certain proportion random lists and valid lists.
4th: the average execution time per list of 10 elements was 0.26 seconds, per list of 100 elements was 1.9 seconds.

(04-21-2017 12:43 AM)DavidM Wrote:  Many of the list processing shortcuts tend to misbehave if you provide them with lists containing no elements, 1 element, or invalid elements. For these challenges, I'm taking you at your word that the input won't include any of those situations.

Yes, I wanted to avoid annoying corner cases. Just a straight processing. I will have to compare your code with mine, your is way more compact. If you want to contribute in the comparison you are welcomed.

Oh, I just understood what your first challenge is doing (compare the nth element with the nth-1 element , I'm not sure about the last element ). That's a beauty! Asap I will compare speed (and results).

Instead someone with a prime may compare the code between han and didier, they do not look the same.