improve solve_models() with tmin/tmax options

[dbrakenhoff]

No description provided.

[dbrakenhoff]

implements partial solution for question in #141.

[codecov]

Codecov Report

Merging #144 into dev will decrease coverage by 0.2%.
The diff coverage is 0%.

@@            Coverage Diff             @@
##              dev     #144      +/-   ##
==========================================
- Coverage   67.35%   67.15%   -0.21%     
==========================================
  Files          31       31              
  Lines        3345     3355      +10     
==========================================
  Hits         2253     2253              
- Misses       1092     1102      +10

Impacted Files	Coverage Δ
pastas/project/project.py	39.28% <0%> (-1.63%)	⬇️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 0e9ad18...1ad658f. Read the comment docs.

[OnnoEbbens]

nice

[raoulcollenteur]

Nice indeed. I think this is a good way of implementing it. Could be made more general for all arguments passen on to ml.solve in the future.

[Hugovdberg]

This could in fact be made more general quite easily by checking the kwargs for pandas.Series objects.

m_kwargs = {}
for key, value in kwargs.items():
    if isinstance(value, pd.Series):
        m_kwargs[key] = value.loc[ml_name]
    else:
        m_kwargs[key] = value
# Convert timestamps
for tstamp in ['tmin', 'tmax']:
    if tstamp in m_kwargs:
        m_kwargs[tstamp] = pd.Timestamp(m_kwargs[tstamp])

[...]
ml.solve(report=report, **m_kwargs)

Also, is there any reason the new arguments are not forced to be passed as keywords (PEP 3102) since Python 3 is required anyway?

[dbrakenhoff]

Good suggestion! We also realized this was something we should eventually make more general, but since I only needed tmin/tmax at the moment, I got lazy.

Feel free to submit a PR with this change, otherwise I'll implement your suggestion sometime later this week.