Add npv_table for net present value of flow-variable changes

[arihantlodha-cmd]

Closes #1131.

Adds output_tables.npv_table, which builds a table of the net present value of the reform-minus-baseline change in flow variables (e.g. Y) over a horizon, evaluated at a list of discount rates. For each variable:

NPV = sum_{t=0}^{num_years-1} (x_reform[t] - x_base[t]) / (1 + r)^t

This follows the pattern the issue points to from the Cost of Disease repo, generalized into a reusable table function that fits the existing output_tables.py conventions (same base_tpi/reform_tpi/*_params signature, VAR_LABELS, and save_return_table for csv/tex/excel/json output).

Design choices, flagged for review:

• Un-stationarized by default. Discounting the stationarized (detrended) values isn't the economically meaningful object, so the variables are un-stationarized before discounting so the NPV runs over the actual trend-inclusive level path. stationarized=True keeps the model values if a caller wants them. (This is why the default differs from macro_table, which defaults to stationarized.)
• Reform required. The table is defined on a reform-minus-baseline change, so reform_tpi is asserted to be present.
• Model units. Results are in the variable's model units. Since Make SS and TPI output dictionaries expose the same keys #1185 added factor to the output dictionaries, a caller can scale by it for dollars; I left that to the caller rather than bake in a dollar-year assumption. Happy to add an optional dollar conversion if you'd prefer it built in.

Output layout: one row per variable, one column per discount rate.

Testing

• test_npv_table builds the table from the cached run output and checks the shape.
• test_npv_table_values checks the discounted sum against a hand calculation at r = 0 and r = 0.1.
• Also verified manually on the cached baseline/reform that NPV falls as the discount rate rises, as expected.
• pytest tests/test_output_tables.py passes (21); ruff format --check ., ruff check ., and linecheck clean.

[codecov-commenter]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 72.85%. Comparing base (a8b0232) to head (4d06c13).

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #1195      +/-   ##
==========================================
+ Coverage   72.74%   72.85%   +0.10%     
==========================================
  Files          22       22              
  Lines        5768     5791      +23     
==========================================
+ Hits         4196     4219      +23     
  Misses       1572     1572              

Flag	Coverage Δ
unittests	72.85% <100.00%> (+0.10%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Files with missing lines	Coverage Δ
ogcore/output_tables.py	88.99% <100.00%> (+0.83%)	⬆️

🚀 New features to boost your workflow:

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