334 locational eligibility based on operation rates

[lovinduw]

When looking into the issue we identified that Transmission capacity min/max/fix or operation min/max/fix can be submitted as Dataframes or as Dataseries. These values are pre-processed in the preprocess2dimData function in Utils. During this pre-processing, if these values are in Dataframe format, only the locations which has a value > 0 are considered. But, if they are in Series format, all the values are considered including 0s. If locationEligibility is not set by the user in the beginning, it will be automatically set such that only the locations which has a value > 0 for capacity min/max/fix or operation min/max/fix are considered as eligible. Later, this locationEligibility is checked with the locations in the processed capacity min/max/fix or operation min/max/fix data and the process fails if they are not the same.

Therefore, if the inputs (capacity min/max/fix) are in Series format with some values 0, this creates error. To avoid this, we added new lines to Utils in preprocess2dimData function and now it only considers values greater than 0 for Series as well (similar to Dataframes) if the LocationEligibility is not set at the beginning.

[phil-fzj]

closes #133

[phil-fzj]

@lovinduw can you check this comment i made last time? #133 (comment)
And if not integrate this?

[JohannesBehrens]

Does this code update include code to set loc elig to 0 when operation rate max is 0 for all time steps/ investment periods?

[JohannesBehrens]

Please add assertion statements to make sure the test works as intended

[JohannesBehrens]

This code looks like it would simply ignore data, which is zero? Doesn't this result in Nones which are ignored later?

[JohannesBehrens]

This becomes a problem when we want to set for example operationRateMax = 0. If this is ignored the operation would not be constrained

[lovinduw]

@JohannesBehrens:

I reviewed the issue from the beginning. The problem occurs when the locationEligibility for transmission components are not explicitly declared and transmission capacityFix = 0 values (for some lines) is explicitly declared. When the locationEligibility is not set, Fine set the locationEligibility based on the capacityFix values. This locationEligibility is set only for the transmission lines with a capacity > 0. For example if capacityMax[loc1_loc2] = 1.5, locationEligibility[loc1_loc2] = 1. If capacityMax[loc1_loc3] = 0, it won't appear in locationEligibility. Therefore, when Fine checks the indexes in the capacityFix and locationEligibility, it creates an error because the number of indexes in capacityFix > number of indexes in locationEligibility. This is because the indexes with capacity 0 in capacityFix are not there in the locationEligibility indexes. Apart from capacityFix, if we define operationMax values with values 0, a similar error is raised. In this case, the indexes in locationEligibility (which is still created based on capacityFix values) are not matching with the columns of operationMax (operationMax is a DataFrame).

However, if only operationMax is defined without defining capacityFix, then the locationEligibility is set for all transmission lines (even for lines with operationMax = 0). Therefore, in this case no error occurs.

Possible solution:

If we can remove the below code lines in transmission.py, which removes the locationEligibility of the lines with capacityFix = 0, the problem will be resolved. But, need your opinion with why we have that code lines in the first place. Is it just to speedup the optimization/ memory save etc.?