Optimization: Replace shift().fillna(0) with shift(fill_value=0) in cumulative time-weight calculation

https://github.com/PyPSA/pypsa-usa/blob/1655e3408b740c4d7ce65d9d951bb39881f6ac4c/workflow/scripts/prepare_network.py#L92
Hi, I’d like to suggest a small performance and clarity improvement in the following line:
start = time_weighting.cumsum().shift().fillna(0)
This can be simplified to:
start = time_weighting.cumsum().shift(fill_value=0)
The shift(fill_value=0) version performs the shift and fill in a single pass, avoiding the creation of an intermediate Series containing NaNs and the overhead of a second pass for fillna(0). Internally, this is implemented more efficiently using lower-level C operations within Pandas.

While functionally identical, the optimized version improves runtime performance and reduces memory allocation, especially when applied in time-critical loops or large-scale pipelines. It also improves code readability by making the default fill behavior explicit in the shift step.

Open workflow/scripts/prepare_network.py at line 92 and inspect the cumulative time-weight calculation. Replace the chained shift().fillna(0) expression with the specified shift(fill_value=0) form, then verify that the calculation remains functionally equivalent. Done means the targeted line uses the single shift call without changing its result.