Unable to distinguish exported curves with packer

[robcalon]

Context

Exporting the curves with the packer doesn't provide a useful dataframe as all types of hourly curves dataframes are concatenated together without a unique key. This makes it impossible to subset specific curves after export.

Steps to reproduce

See the following notebook snippet and resulting dataframe.

from pyetm.models import Session, Scenarios, ScenarioPacker

s1 = Session.load(1172446)
s2 = Session.load(1172447)
s3 = Session.load(1172450)

scenarios = Scenarios(items=[s1, s2, s3])

packer = ScenarioPacker()
packer.add(*scenarios)

exports = packer.exports()
exports

Proposed solution

Please add a (named) level to the multi index column that specifies the type of curve so that this curve type can be extracted. It would also be very useful if the curve type(s) to be exported can be added as an optional parameter in the exports call itself, e.g. packer.exports(curves=["electricity", "electricity_price"]). Furthermore you could consider offering the exported curves as a dictionary of dataframes instead of concatenating them all together into a single dataframe.

[robcalon]

@noracato to continue the conversation via email in this thread. Would you see the logic in implementing the exported curves as a dictionary as well? This would already mostly address the request in #133 to make more custom exports possible. Thereafter it also seems more consistent with the logic implemented in the Session and Scenario object's itself (also a dictionary).