feat: signal generator — weighted ensemble with market data

services/signal_generator/market_data.py

@@ -0,0 +1,122 @@
+"""In-memory market data manager with rolling OHLCV windows.
+
+Maintains a per-ticker deque of recent bars and computes technical
+indicators (SMA, RSI) on demand when building ``MarketSnapshot`` objects.
+"""
+
+from __future__ import annotations
+
+from collections import deque
+from typing import Any
+
+from shared.schemas.trading import MarketSnapshot, OHLCVBar
+
+
+# Default rolling-window sizes
+_DEFAULT_MAX_BARS = 100
+_RSI_PERIOD = 14
+
+
+class MarketDataManager:
+    """Manages in-memory rolling windows of OHLCV bars per ticker.
+
+    Parameters
+    ----------
+    max_bars:
+        Maximum number of bars to retain per ticker.
+    """
+
+    def __init__(self, max_bars: int = _DEFAULT_MAX_BARS) -> None:
+        self.max_bars = max_bars
+        self._bars: dict[str, deque[OHLCVBar]] = {}
+
+    # ------------------------------------------------------------------
+    # Public API
+    # ------------------------------------------------------------------
+
+    def add_bar(self, ticker: str, bar_data: dict[str, Any] | OHLCVBar) -> None:
+        """Append a bar to the rolling window for *ticker*.
+
+        ``bar_data`` can be a dict (parsed from JSON) or an ``OHLCVBar``
+        instance.
+        """
+        if isinstance(bar_data, dict):
+            bar = OHLCVBar.model_validate(bar_data)
+        else:
+            bar = bar_data
+
+        if ticker not in self._bars:
+            self._bars[ticker] = deque(maxlen=self.max_bars)
+        self._bars[ticker].append(bar)
+
+    def get_snapshot(self, ticker: str) -> MarketSnapshot | None:
+        """Build a ``MarketSnapshot`` from the rolling window.
+
+        Returns ``None`` if no bars have been recorded for *ticker*.
+        """
+        bars = self._bars.get(ticker)
+        if not bars:
+            return None
+
+        latest = bars[-1]
+        closes = [b.close for b in bars]
+
+        return MarketSnapshot(
+            ticker=ticker,
+            current_price=latest.close,
+            open=latest.open,
+            high=latest.high,
+            low=latest.low,
+            close=latest.close,
+            volume=latest.volume,
+            sma_20=self._compute_sma(closes, 20),
+            sma_50=self._compute_sma(closes, 50),
+            rsi=self._compute_rsi(closes, _RSI_PERIOD),
+            bars=[b.model_dump(mode="json") for b in bars],
+        )
+
+    def has_ticker(self, ticker: str) -> bool:
+        """Return ``True`` if at least one bar exists for *ticker*."""
+        return ticker in self._bars and len(self._bars[ticker]) > 0
+
+    # ------------------------------------------------------------------
+    # Technical indicator helpers
+    # ------------------------------------------------------------------
+
+    @staticmethod
+    def _compute_sma(closes: list[float], period: int) -> float | None:
+        """Compute the simple moving average over the last *period* closes.
+
+        Returns ``None`` if there are fewer than *period* data points.
+        """
+        if len(closes) < period:
+            return None
+        return sum(closes[-period:]) / period
+
+    @staticmethod
+    def _compute_rsi(closes: list[float], period: int = 14) -> float | None:
+        """Compute the standard RSI over the last *period+1* closes.
+
+        Uses the average-gain / average-loss method.  Returns ``None`` if
+        there are not enough data points (need at least ``period + 1``
+        closes to compute ``period`` deltas).
+        """
+        if len(closes) < period + 1:
+            return None
+
+        # Only use the most recent period+1 closes
+        relevant = closes[-(period + 1):]
+        deltas = [relevant[i + 1] - relevant[i] for i in range(len(relevant) - 1)]
+
+        gains = [d for d in deltas if d > 0]
+        losses = [-d for d in deltas if d < 0]
+
+        avg_gain = sum(gains) / period if gains else 0.0
+        avg_loss = sum(losses) / period if losses else 0.0
+
+        if avg_loss == 0:
+            return 100.0  # No losses -> RSI is 100
+
+        rs = avg_gain / avg_loss
+        rsi = 100.0 - (100.0 / (1.0 + rs))
+        return round(rsi, 4)