Partitioning a list by using relative differences

Replies to This Discussion

Permalink Reply by Behnood Eghbali on March 27, 2013 at 2:43am

I wrote a script to get the size of partitions but, unfortunately it didn't work.

Permalink Reply by Jesus Galvez on March 27, 2013 at 3:04am

The size of the list of your [Partition List] component is:

1, 3, 5, 1, 4, ...

What's the logic of that list?

Permalink Reply by Behnood Eghbali on March 27, 2013 at 3:16am

True, the logic comes from relative differences between sorted list values.

Permalink Reply by phillip on March 27, 2013 at 3:21am

This is a solution for a similar problem I had once. Not exactly what you want and also in times before PartitionList, but you might be able to adapt. 

Attachments:

• overlapdomain pcr.gh, 6 KB

Permalink Reply by Behnood Eghbali on March 27, 2013 at 3:35am

Thank you phillip, nice! so how can I reach to the values in each domain?

Permalink Reply by Chris Tietjen on March 27, 2013 at 5:46am

Use the Successive Difference component.

Permalink Reply by phillip on March 27, 2013 at 7:49am

You are right, Chris. Successive Difference and Partition List are the best components here. But I think this is more what Behnood is after:

Two comments though:

- Would be nice if [Member Index] would accept the String right away

- How about a Partition by Index component?

Attachments:

• relativePartition pcr.gh, 4 KB

Permalink Reply by Behnood Eghbali on March 27, 2013 at 8:18am

Bingo! :) Thank you so much phillip! That's what I'm talking about. :) Cheers!

Permalink Reply by Chris Tietjen on March 27, 2013 at 9:42am

Did I get it this time or am I still missing something?

Permalink Reply by Behnood Eghbali on March 27, 2013 at 9:52am

Comparing each 2 values or numbers together. Expected output for this one is: {0}:1,2 ; {1}: 5,6 ; {2}: 9 ; {3}: 12,13 ; {4}: 15,16 ...

Permalink Reply by Chris Tietjen on March 27, 2013 at 2:53pm

I get it. Thanks for explaining.

Permalink Reply by Behnood Eghbali on March 27, 2013 at 8:12am

Thank you Chris, I tried it, but something was missing. I want to group some random sorted data just like the screenshot I uploaded above.