Version Checks (indicate both or one)

• I have confirmed this bug exists on the lastest release of PyPSA.

• I have confirmed this bug exists on the current master branch of PyPSA.

Issue Description

get_scenario() leaves MultiIndex on empty components, which breaks n.optimize() call.

The problem is specifically here:

n['low'] yields the same problem, so __getitem__ has the same issue.

I will patch both via PR in a moment.

Reproducible Example

import pypsa

n = pypsa.Network()
n.set_snapshots(range(3))
n.add("Bus", "bus")
n.add("Load", "load", bus="bus", p_set=100)
n.add("Generator", "gen", bus="bus", p_nom_extendable=True, capital_cost=1000)

n.set_scenarios({"low": 0.5, "high": 0.5})

n_low = n.get_scenario("low")  # or n["low"]
n_low.optimize(solver_name="highs")  # ValueError: conflicting multi-index level name 'name'

print(n_low.has_scenarios) # False
print(n_low.c.links.static.index) # MultiIndex([], names=['scenario', 'name'])

Expected Behavior

Methods should reset MultiIndex and n.optimize() call should work on retrieved networks.

Installed Versions

v1.0