Add explicit flag for categorical parameters to ARES/PyARES

[arthursloan]

Right now when creating a parameter for an ARES OS, you select the data type for the parameter. Some of these data types (e.g., booleans, strings) are effectively inherently categorical, however even numeric values could be categorical under some circumstances. When passing parameters to a planner that supports categorical variables it will need to be informed which variables are part of a continuous range and which are categorical. It would be useful to have a flag the user can set to mark a parameter as categorical that is passed along to any planners as a parameter_type field just like minimum_value, maximum_value, or data_type

[arthursloan]

I'll add on to this, we should also distinguish between ordinal choices, where there is an inherent order (e.g., low, medium, high) and truly categorical variables (e.g., red, green, blue, yellow, purple). This would probably take the form of an "is ordered" flag that accompanies an "is categorical" flag.