Fix curves finally

[lsbardel]

No description provided.

[Copilot]

Pull request overview

This PR standardizes how volatility-surface forwards and discount curves are calibrated from put-call parity, removes user-selectable curve models from the API/UI, and introduces a consistent ref_date for time-to-maturity calculations across loaders, surfaces, and calibration.

Changes:

• Make ref_date an explicit field on surface pricers/loaders and propagate it into VolSurface.
• Rework curve calibration flow: asset curve is always fitted from parity forwards, quote curve is fitted when calibratable (otherwise treated as known), and API/UI curve selection is removed.
• Simplify parity tooling by removing discount-factor fitting helpers and updating tests/docs/examples to the new calibration approach.

Reviewed changes

Copilot reviewed 15 out of 15 changed files in this pull request and generated 2 comments.

Show a summary per file

File	Description
quantflow/rates/interpolated.py	Use shared rounding precision (ROUND_RATE) when setting interpolated curve anchor rates during calibration.
quantflow/options/surface.py	Introduce explicit ref_date, change default asset curve model, refactor curve calibration behavior, and update surface construction.
quantflow/options/parity.py	Remove discount-factor fitting utilities and adjust plotting to use forward + quote-discount estimation.
quantflow/data/yahoo.py	Initialize loaders with explicit ref_date and use interpolated curves as the starting quote curve model for equities.
quantflow/data/deribit.py	Initialize loaders with explicit ref_date and rely on default curve setup appropriate for Deribit flows.
quantflow_tests/test_surface_calibration.py	Update calibration tests to match new curve-fitting semantics and remove tests for deleted helpers.
quantflow_tests/test_parity.py	Update parity tests to validate the new quote-discount estimation approach.
quantflow_tests/test_non_inverse_surface.py	Ensure loaders are built with an explicit ref_date.
quantflow_tests/test_app.py	Update API behavior assertions now that curve selection is backend-driven and cached by asset only.
frontend/src/volatility-surface.md	Remove curve selectors and align UI copy with the new backend-driven curve calibration behavior.
docs/tutorials/volatility_surface.md	Update tutorial narrative and references to the new forward/curve calibration flow.
docs/tutorials/curve_calibration.md	Expand and clarify the curve calibration derivation around fixed forwards and quote discount estimation.
docs/examples/curve_calibration.py	Update example workflow to use calibrate_forwards outputs and simplify dataframe construction.
app/api/volatility.py	Remove curve selection query params and always run backend-driven calibrate_curves().
app/api/docs/volatility_surface.md	Align API documentation with the new fixed curve calibration behavior.

Suppressed comments (2)

quantflow/options/surface.py:1567

• The calibrate_curves() docstring says that when the asset curve model cannot be calibrated (e.g. NoDiscountCurve) an InterpolatedMonotonicCubicCurve is fitted instead, but the implementation now raises ValueError in that case. Please update the docstring to reflect the actual behavior (or implement the fallback).

        cannot be calibrated, for example a NoDiscountCurve, an
        [InterpolatedMonotonicCubicCurve][quantflow.rates.interpolated.InterpolatedMonotonicCubicCurve]
        is fitted instead: it passes exactly through its nodes, so the curve
        implied forward reproduces the parity forward at every calibrated
        maturity.

quantflow/options/surface.py:1522

• The asset_curve parameter docs say that passing None keeps the current curve and treats it as known, but the implementation always fits the asset curve from parity forwards (and raises if the curve has no calibrator). The parameter docs should reflect this behavior.

                "YieldCurve type or instance to fit the asset discount curve "
                "$D_a$. "
                "When None the current asset_curve is kept and treated as known."
            ),

[codecov-commenter]

Codecov Report

❌ Patch coverage is 64.70588% with 6 lines in your changes missing coverage. Please review.
✅ Project coverage is 89.03%. Comparing base (5e16f37) to head (a203dd1).
⚠️ Report is 8 commits behind head on main.

Files with missing lines	Patch %	Lines
quantflow/options/parity.py	16.66%	5 Missing ⚠️
quantflow/options/surface.py	80.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #92      +/-   ##
==========================================
+ Coverage   88.75%   89.03%   +0.27%     
==========================================
  Files          84       85       +1     
  Lines        5142     5410     +268     
==========================================
+ Hits         4564     4817     +253     
- Misses        578      593      +15     

☔ View full report in Codecov by Harness.
📢 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.