Rename SymbolicPermutation...? about complexitymeasures.jl HOT 4 CLOSED

Why isn't it just called OrdinalPatterns and WeightedOrdinalPatterns?

I wrote my answer too quickly: I like your suggestions better. OrdinalPatterns, WeightedOrdinalPatterns and AmplitudeAwareOrdinalPatterns.

EDIT: However, skipping "patterns" makes the name shorted, and "ordinal" already implies some sort of ordering pattern, so perhaps the shorter alternatives proposed above work too. I'm not too attached to any of the names, as long as we drop the "Symbolic" part in favor of "Ordinal".

from complexitymeasures.jl.

me2

from complexitymeasures.jl.

I never quite understood why we are using different words to refer to the same thing. We have "ordinal pattern encoding" but the outcome space is called "symbolic permutation". Why? Why isn't it just called OrdinalPatterns and WeightedOrdinalPatterns?
Wouldn't this simplify things a lot? Alternative is to rename the encoding to be SymbolicPermutationEncoding. But I see no reason to use two different names for the same thing.

We went a bit back and forth on the naming, and I agree that we might have ended on a sub-par alternative. The Symbolic prefix was there because it originated from the permutation entropy. But since SymbolicPermutation and friends now are outcome spaces, there's no reason to keep this name. Moreover, every discretization scheme "symbolizes" - that is what happens when you discretise. I think Ordinal-something is much more informative. I suggest the following outcome space renamings:

• SymbolicPermutation -> Ordinal
• SymbolicAmplitudeAwarePermutation -> OrdinalAmplitudeAware
• SymbolicWeightedPermutation -> OrdinalWeighted

The OrdinalPatternEncoding we can either keep, or rename to OrdinalEncoding (btw: there will soon be OrdinalAmplitudeAwareEncoding and OrdinalWeightedEncoding too, which will allow us to explicitly use these outcome spaces in upstream functions such as transfer entropy).

from complexitymeasures.jl.

After my edit above:

1. Ordinal/OrdinalWeighted/OrdinalAmplitudeAware
2. OrdinalPatterns/WeightedOrdinalPatterns/AmplitudeAwareOrdinalPatterns

I vote for alternative 2. It's more intuitive.

from complexitymeasures.jl.