Correct set alignment

[trevorb1]

In this PR I add functionality to define option data over any set. for example, to apply an emission limit and a capacity constraint to the same scenario data, you can define data like so:

PARAMETER,REGION,TECHNOLOGY,EMISSION,OPTION,YEAR,VALUE
TotalAnnualMaxCapacityInvestment,UTOPIA,IMPHCO1,,0,1990,999999999
TotalAnnualMaxCapacityInvestment,UTOPIA,IMPHCO1,,0,1991,999999999
TotalAnnualMaxCapacityInvestment,UTOPIA,IMPHCO1,,0,1992,999999999
...
AnnualEmissionLimit,UTOPIA,,NOX,0,,-50 
AnnualEmissionLimit,UTOPIA,,NOX,1,,-10000

While this bypasses the error described in issue #85, I am a little confused on how logic of handling data that is not defined over all sets (ie. AnnualEmissionLimit does not have an associated year in the provided example in the issue ticket). @HauHe was this just a mistake? Im not sure where to look to see if AnnualEmissionLimit is being applied correctly. I would suggest explicitly defining the year.

[HauHe]

Thanks a lot @trevorb1!
I will have a look at it tomorrow.
If I remember correctly I had left the year here on purpose, but I'm not 100% sure anymore what behaviour I expected when applying the data. Has been some time since I wrote the code. I'll let you know

[HauHe]

Hi @trevorb1 ,
sorry for the long silence, was a bit difficult to find the time to work on this.
In the original version of OSeMOSYS_step, I had written some code that allowed to instead of providing a series of values for each year in the options, to provide a single value that could be either absolute or a percentage, indicated via the %-symbol. This only worked if the parameter was already defined in the step before, if not an error was produced. I think this feature is very useful since it allows to have changes in parameters over the modelling time horizon without having jumps, since you are indicating the change of the parameter an not the absolute value.
I think it would be great if we could reintroduce this feature.