Got it, thanks!

OK... I love this kata, and I've got it mostly working using a k-means algorithm with a solution specific centroid initializer in C#. I am getting an odd issue with the testShortMessages input bit string of 101.
To troubleshoot this, just before returning my dot-dash sequence, I override my output on the input of 101:

• If I set the output to equal to EE (. .) I get an error saying it should be I
• If I set the output to equal to I (..) I get an error saying it should be EE

Any which way I solve this, it appears I will always fail one of these test cases... any thoughts?

Awsome Kata!!! I just wish I could submit it as I am also getting the "763 but was 872" error.

It is indeed a clever solution and may be a best practice for “shops” that use C#/LINQ. However, if you have a “shop” with diverse developer backgrounds, or you need to port code to other languages on occasion, Classic C# (no LINQ) would probably be a best practice.