feat: add RMA (Running Moving Average / Wilder's Smoothing) indicator

[mac-agent]

Summary

Adds the RMA (Running Moving Average) indicator, implementing the equivalent of TradingView Pine Script ta.rma().

Fixes #141.

What is RMA?

RMA is also known as Wilder's Smoothing Average. It's used extensively in technical analysis — most notably in RSI calculations, ADX, and other Wilder-derived indicators.

Formula: RMA is equivalent to an EMA with alpha = 1/period:

• First value: SMA of first N values
• Subsequent: RMA = (prev_RMA * (period - 1) + current) / period

Implementation

• Uses pandas ewm(alpha=1/period, adjust=False, min_periods=period).mean() for efficient vectorized computation
• Follows existing function signature pattern (dataframe, period, field)
• Added to overlap_studies.py alongside other moving averages
• Exported via existing wildcard imports

Testing

Added test_rma with 3 assertions:

• Produces non-null values after warmup period
• All values positive for BTC price data
• Less volatile than raw prices (smoothing property)

[xmatthias]

I'd want to have a test in test_indicators_generic for this, too.

That test ensures the interface is aligned with the others. You'll have to refresh the snapshots - but that's no issue, really.

We'll also want to add it to the index / docs page - otherwise it'll be code that nobody knows about.