search

NicMcPhee / lexicase-clusturing-analysis

Scripts and tools (but not data!) for doing clustering analysis to try to better understand lexicase selection.
MIT License
0 stars 1 forks source link

Sort RSWN test cases #1

Open NicMcPhee opened 9 years ago

NicMcPhee commented 9 years ago

The even test cases are totally different than the odd test cases, and we'd like to renumber them so the evens are together and the odds are together. In fact there are other problems with 3 different values per test case, so we want to update the code to take the number of values per test case as a parameter and renumber the test cases (probably as part of the scaling) so that the related test cases are together.

NicMcPhee commented 9 years ago

So I think we have the pair of values case sorted out (commit 0baf69b3fe27c4577f3f800d45376d1a8724bf43), but it's currently fixed on exactly two values. Thus it will break if there's just one, and won't generalize to three. That's all do-able, it's just not done.