I am trying to convert this notebook on probability by Peter Norvig . I observe that Python runs much faster than Julia. I couldn’t pin down the problem in my Julia code. What is happening here is that we are looping over a massive array of strings that represent all the possible combinations of 5 c…

I thought it is worth to elaborate more on what @dpsanders proposed. Apart from using Set in the computations I think that the problem is that four_kind(hand)=any(count_o(hand,rank) == 4 for rank in ranks) uses ranks that is a global variable. Using local variable in your definition of four_kind …

You are of course completely correct. Embarrassing that I didn’t notice the first error that we always tell people to avoid!! Try wrapping everything in a function .

Thanks @dpsanders and @bkamins So this the latest version I have so far using Combinatorics "counts the number of occurances of l in s " count_o(s,l)=count(x->x==l,s) "The set of ways of concatenating one item from collection A with one from B." cross(A, B)=["$a$b" for a in A for b in B] "All com…

For the sake of completeness, the implementation using sets is using Combinatorics "counts the number of occurances of l in s " count_o(s,l)=count(x->x==l,s) "The set of ways of concatenating one item from collection A with one from B." cross(A, B)=["$a$b" for a in A for b in B] "All combinations…

I continued with more experiments where I tried to preserve use of Sets so as not to deviate from the python code. In the first variation, I used collect instead of simply passing the generator to the Set constructor. function combos(items, n) Set(collect(join(combo," ") for combo = combinatio…

On the risk of getting off topic, I would like to encourage a non-string solution for this problem. I see two problems with the string approach: The current implementation relies on the fact that the characters in suits and ranks don’t overlap, which is just a fortuitousness in the English languag…

Probably some more speed can be gained (also in the original version) by replacing the flush function to just check if all the suits are equal to each other.

@dpsanders is right, in my version replacing flush by function flush(hand) s1 = hand[1].suit count(card -> card.suit==s1, hand) == 5 end yields 0.388377 seconds (15.59 M allocations: 872.454 MB, 11.35% gc time) i.e. another ~25% faster.

Instead of an enum, you can just use integers instead of Chars for suits and ranks, and get a similar speedup. I am curious as to why there’s still so much allocation.

Fast String processing in Julia vs. Python

General Usage

StefanKarpinski March 7, 2017, 2:51pm 39

I’ve split the discussion of my new Cards.jl package into a new thread.

Topic		Replies	Views
Optimizing performance (newcomer from Matlab) General Usage first-steps	19	1396	November 8, 2019
Julia slower than Python for finding pangrams? Performance question	28	2048	April 5, 2024
In as few lines as possible describe why you love julia New to Julia	113	13463	May 21, 2021
General questions from Python user Performance	59	4846	March 8, 2021
Quite bad performance of Julia 0.6.4 vs Python+Numpy General Usage	26	5536	November 13, 2018

Fast String processing in Julia vs. Python

Related topics