Reach a variable number with the sum of fixed numbers

Replies to This Discussion

Permalink Reply by Anders Holden Deleuran on November 10, 2017 at 1:48am

One method that springs to mind would be combinatorial generation. That is, for a list of values, generate all combinations/permutations of these. There's a thread on various options in GH here (I'd personally recommend the Python itertools one). So in your case, for instance: generate all combinations (using the second argument to generate only sets of three), calculate the sum of all the generated sets, and then find the sum closest to X.

Permalink Reply by Anders Holden Deleuran on November 10, 2017 at 2:04am

Here's an quick go (didn't test, so might be buggy, looks okay though):

Edit: You might expand this to start the search with the SetSize at 1, check if this generates a result that is within some tolerance distance from X, if not, jump to 2, if not, jump to 3 etc. A less exhaustive/brute force method might be to implement Galapagos for the search, but with smaller sets I suspect that won't be an advantage (also, it comes with its own can of worms). That said, the search space and amount of possible combinations grows rapidly with these cases.

Attachments:

• 171110_FindClosestSetSumCombination_GHPython_Itertools_00.gh, 7 KB

Permalink Reply by Petrik Kollár on November 10, 2017 at 6:16am

I used your python script and  I'm getting closer to the solution,

Here is where i'm standing now.

I set 79 as goal, my result is 81 which is good, but not the best optimalization,

Becuase it gives me  3pcs of 12   and   5 pcs of 9

the good solution would be  4pcs of 18 and 1pcs of 9

Attachments:

• reach.gh, 17 KB

Permalink Reply by Anders Holden Deleuran on November 10, 2017 at 6:29am

In that case, you might have more succes using the function combinations_with_replacement():

Edit: Also note that the script is written assuming floats, you might need to edit it for integer logics (i.e. floor/ceiling type pitfalls etc.).

Permalink Reply by Petrik Kollár on November 10, 2017 at 7:00am

i don't understand pyhton unfortunately, where can i edit floats to integer? 

Permalink Reply by Anders Holden Deleuran on November 10, 2017 at 7:50am

In the GHPython component script editor (double-click one and it opens up, like in the screenshots).

Permalink Reply by RAF on November 19, 2017 at 1:31pm

Hello
I am wondering if its possible to ask a question here
Is their something in script
That may add another input to "mass addition"
1- the value
2- the count that mass addition operates?
Isnt it possible ?
I mean something close to Log n (x)

Thank you

Permalink Reply by David Rutten on November 19, 2017 at 6:14pm

There's a Logarithm component. I don't think I understand the question though, perhaps better if you create a new topic.

Permalink Reply by RAF on November 21, 2017 at 4:11am

exactly my question isn't clear enough,

i will formulate differently.

Thank you

Permalink Reply by RAF on November 21, 2017 at 4:25am

i am sorry to stay on for this

i tried the expression editor

Log(x[,b])

i have the 1519 invalid expression error message

Attachments:

• Expression_Log(x[,b]).jpg, 12 KB
• Expression_Log(x[,b])_error1519.jpg, 116 KB

Permalink Reply by Tim Fischer on November 15, 2018 at 6:33am

Hey, I ran into the same error and found a solution.

So for anyone who still has this issue:

Use: Log(x,(CDbl(y))) instead of Log(x[,y])

for "x,y" input

(works also for other formats than sliders as inputs;)

Best Regards

Tim

Attachments:

• Canvas at 14;12;09.png, 176 KB

Permalink Reply by Anders Holden Deleuran on November 20, 2017 at 3:07am

A bit lost as well, there are a couple of logarithmic functions in the Python math module as well (if that was the question).

• ‹ Previous
• 1
• 2
• Next ›
• Page