Add Yang Zhang Volatility (YZV) indicator

[claygeo]

Summary

Implements the Yang Zhang (2000) drift-independent volatility estimator, the most efficient known estimator of historical volatility for financial time series. Closes #8145.

New indicator: YangZhangVolatility combines three variance components with an optimal weighting factor:

• Overnight variance σ²_o = Var(ln(Open_i / Close_{i-1}))
• Intraday variance σ²_c = Var(ln(Close_i / Open_i))
• Rogers-Satchell variance σ²_RS = (1/n) Σ[ln(H/C)·ln(H/O) + ln(L/C)·ln(L/O)]
• Combined: σ²_YZ = σ²_o + k·σ²_c + (1-k)·σ²_RS where k = 0.34 / (1.34 + (n+1)/(n-1))

Files changed:

• Indicators/YangZhangVolatility.cs — New indicator extending BarIndicator, IIndicatorWarmUpPeriodProvider. Uses 5 internal Sum sub-indicators for O(1) rolling computation. Guards zero/negative prices and uninitialized state.
• Algorithm/QCAlgorithm.Indicators.cs — YZV() helper method (same pattern as RSV())
• Tests/Indicators/YangZhangVolatilityTests.cs — Extends CommonIndicatorTests<IBaseDataBar> with hand-computed verification test (YzvComputesCorrectly)
• Tests/TestData/spy_with_yzv.csv — 500-row SPY reference data generated with independent Python implementation

Design decisions

• WarmUpPeriod = period + 1: First bar is a seed (stores _previousClose), computation starts on bar 2, IsReady at bar period+1.
• Sample variance for overnight/intraday, population mean for RS: Matches the Yang Zhang (2000) paper formulation.
• Math.Max(0m, yzVariance) before sqrt: Prevents NaN from floating-point rounding. Matches Variance.cs convention.
• <= 0 guards on all OHLC prices and _previousClose: Prevents Math.Log of zero/negative values producing Infinity/NaN.
• Period >= 2 validation: Constructor throws ArgumentOutOfRangeException to prevent divide-by-zero in (period - 1).

Test Coverage

All 8 code paths tested (100%):

• Constructor validation, seed bar, _previousClose guard, not-ready state, full computation, rolling window (500 bars via CSV), and Reset.

Test plan

• All NUnit tests pass (CommonIndicatorTests suite + YzvComputesCorrectly)
• CSV reference data matches Python-generated expected values
• Edge cases: zero prices, negative prices, period=2, reset mid-stream

Plan Completion

13/13 plan items DONE (1 CHANGED: <= 0 guards instead of == 0). All files created, all specifications met.

Reference

Yang, D. and Zhang, Q. (2000). "Drift Independent Volatility Estimation Based on High, Low, Open, and Close Prices." Journal of Business, 73(3), 477-491.

[Martin-Molinero]

Closing for now until previous PR is approved and merged #9388