Make pandas, scipy, and matplotlib optional dependencies via attempt_import

[Copilot]

Several packages used by mpi-sppy (pandas, scipy, matplotlib) were either undeclared or treated as hard requirements, preventing lightweight installs for users who don't need those features.

Changes

Package metadata (pyproject.toml, setup.py)

• Removed pandas and scipy from required dependencies
• Added optional extras: [pandas], [scipy], [plot] (matplotlib)

Deferred imports — replaced direct imports with Pyomo's lazy-loading mechanism:

• pandas: 4 files (rho_utils, wtracker, phtracker, primal_dual_converger)
• scipy: 5 files (prox_approx, seqsampling, zhat4xhat, mmw_ci, kkt/interface)
• matplotlib.pyplot: 2 files (phtracker, primal_dual_converger)

# Before
import pandas as pd
import scipy.stats
from scipy.optimize import least_squares
import matplotlib.pyplot as plt

# After
from pyomo.common.dependencies import attempt_import, pandas as pd, scipy
plt, plt_available = attempt_import('matplotlib.pyplot')
# Usage: scipy.stats.t.ppf(...), scipy.optimize.least_squares(...)

The deferred import pattern (already used in the repo for networkx, dill, yaml) allows modules to load without these packages installed; errors are raised only when the optional functionality is actually invoked.

[bknueven]

@DLWoodruff I'm a bit ambivalent about having scipy as a soft dependency. We don't use it a lot, so it some ways it's not a huge issue. What do you think?

[codecov]

Codecov Report

❌ Patch coverage is 92.30769% with 1 line in your changes missing coverage. Please review.
✅ Project coverage is 70.17%. Comparing base (e496071) to head (cdc5927).
⚠️ Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
mpisppy/confidence_intervals/zhat4xhat.py	0.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #646      +/-   ##
==========================================
+ Coverage   70.15%   70.17%   +0.02%     
==========================================
  Files         153      153              
  Lines       18905    18905              
==========================================
+ Hits        13262    13267       +5     
+ Misses       5643     5638       -5     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[DLWoodruff]

@DLWoodruff I'm a bit ambivalent about having scipy as a soft dependency. We don't use it a lot, so it some ways it's not a huge issue. What do you think?

As we do more with the stochastic problem itself (e.g., confidence intervals), we will use it a little more; however, it is hard to argue that a soft dependency really hurts anyone. It makes some sense to have as many dependencies as possible be soft. So I say we merge this.