[huuan26]

Resolves #180. Resolves #29.

Summary

This PR adds support for Weighted Moving Average (WMA) and Hull Moving Average (HMA) as first-class indicator columns in StockDataFrame.

Features

1. Weighted Moving Average (WMA):

   • Exposes df['wma'], df['close_10_wma'], df['wma_20'], etc.
   • Default window: 10 on close.
   • Adds class method StockDataFrame.wma(series, window) aliasing the vectorized linear_wma.

2. Hull Moving Average (HMA):

   • Implements Alan Hull's zero-lag formulation:
     • $W_1 = \text{WMA}(p, \lfloor n/2 \rfloor)$
     • $W_2 = \text{WMA}(p, n)$
     • $\text{Diff} = 2 W_1 - W_2$
     • $\text{HMA} = \text{WMA}(\text{Diff}, \lfloor\sqrt{n}\rfloor)$
   • Exposes df['hma'], df['close_16_hma'], df['hma_9'], etc.
   • Default window: 16 on close.
   • Handles edge cases cleanly ($window = 1$, $window > n$, zeroed warmup period).

3. Handler Key Normalization:

   • Stores default window entries as 1-tuples in self.handler to ensure exact membership lookup and prevent substring collisions (such as 'wma' matching 'vwma').

Verification & Documentation

• Added unit tests in test.py:
  • test_wma: Verification against 30-day stock data and default/custom windows.
  • test_hma: Verification against 90-day stock data, warmup handling, and custom windows.
  • test_hma_edge_cases: Edge cases ($window=1$, $window > n$, step-by-step benchmark verification).
• Updated README.md Summary table and Moving Averages section with formulas and usage examples.
• Code style: flake8 stockstats.py test.py passed with 0 errors.
• Test suite: 131/131 tests passed.