Implement M3 projected daily moving averages

app/analysis/__init__.py

@@ -0,0 +1 @@
+"""Pure analysis engines."""

app/analysis/indicators.py

@@ -0,0 +1,40 @@
+from app.bars.models import Bar
+
+
+def sma(values: list[float], period: int) -> list[float | None]:
+    if period <= 0:
+        raise ValueError("period must be positive")
+    output: list[float | None] = [None] * len(values)
+    total = 0.0
+    for index, value in enumerate(values):
+        total += value
+        if index >= period:
+            total -= values[index - period]
+        if index >= period - 1:
+            output[index] = total / period
+    return output
+
+
+def ema(values: list[float], period: int) -> list[float | None]:
+    if period <= 0:
+        raise ValueError("period must be positive")
+    output: list[float | None] = [None] * len(values)
+    if len(values) < period:
+        return output
+    value = sum(values[:period]) / period
+    output[period - 1] = value
+    multiplier = 2 / (period + 1)
+    for index in range(period, len(values)):
+        value = (values[index] - value) * multiplier + value
+        output[index] = value
+    return output
+
+
+def atr(bars: list[Bar], period: int = 14) -> list[float | None]:
+    if period <= 0:
+        raise ValueError("period must be positive")
+    ranges: list[float] = []
+    for index, bar in enumerate(bars):
+        previous_close = bars[index - 1].c if index else bar.c
+        ranges.append(max(bar.h - bar.l, abs(bar.h - previous_close), abs(bar.l - previous_close)))
+    return sma(ranges, period)

app/analysis/levels.py

@@ -0,0 +1,48 @@
+from dataclasses import asdict, dataclass
+from enum import Enum
+from typing import Any
+
+from app.bars.models import Timeframe
+
+
+class LevelKind(str, Enum):
+    MANUAL = "manual"
+    MA = "ma"
+    TRENDLINE = "trendline"
+    HORIZONTAL = "horizontal"
+
+
+class Side(str, Enum):
+    SUPPORT = "support"
+    RESISTANCE = "resistance"
+
+
+@dataclass(slots=True)
+class Level:
+    id: str
+    kind: LevelKind
+    tf: Timeframe
+    side: Side
+    weight: float
+    score: float
+    label: str
+    anchor_t: int
+    anchor_p: float
+    slope: float
+    points: list[tuple[int, float]] | None
+    touches: int
+    first_t: int
+    last_t: int
+    provisional: bool
+    hidden: bool
+    period: int | None = None
+
+    def price_at(self, t: int) -> float:
+        return self.anchor_p + self.slope * (t - self.anchor_t)
+
+    def to_dict(self) -> dict[str, Any]:
+        value = asdict(self)
+        value["kind"] = self.kind.value
+        value["tf"] = self.tf.value
+        value["side"] = self.side.value
+        return value

app/analysis/moving_averages.py

@@ -0,0 +1,62 @@
+from app.analysis.indicators import ema, sma
+from app.analysis.levels import Level, LevelKind, Side
+from app.bars.models import Bar, Timeframe
+from app.config import MA_WEIGHT_FACTOR, TIMEFRAME_WEIGHT
+
+
+MA_FUNCTIONS = {"sma": sma, "ema": ema}
+
+
+def build_ma_levels(
+    bars_by_tf: dict[Timeframe, list[Bar]],
+    ma_sets: dict[Timeframe, list[tuple[str, int]]],
+) -> list[Level]:
+    levels: list[Level] = []
+    for tf, definitions in ma_sets.items():
+        bars = bars_by_tf.get(tf, [])
+        closed_count = sum(bar.closed for bar in bars)
+        closes = [bar.c for bar in bars]
+        for kind, period in definitions:
+            if closed_count < period or kind not in MA_FUNCTIONS:
+                continue
+            values = MA_FUNCTIONS[kind](closes, period)
+            points = [(bar.t, value) for bar, value in zip(bars, values) if value is not None]
+            if not points:
+                continue
+            current = points[-1][1]
+            provisional = not bars[-1].closed
+            levels.append(
+                Level(
+                    id=f"ma:{tf.value}:{kind}:{period}",
+                    kind=LevelKind.MA,
+                    tf=tf,
+                    side=Side.SUPPORT if current <= bars[-1].c else Side.RESISTANCE,
+                    weight=TIMEFRAME_WEIGHT[tf] * MA_WEIGHT_FACTOR,
+                    score=1.0,
+                    label=f"{tf.value} {kind.upper()}{period}",
+                    anchor_t=points[-1][0],
+                    anchor_p=current,
+                    slope=0.0,
+                    points=points,
+                    touches=0,
+                    first_t=points[0][0],
+                    last_t=points[-1][0],
+                    provisional=provisional,
+                    hidden=False,
+                    period=period,
+                )
+            )
+    return levels
+
+
+def project_step(points: list[tuple[int, float]], bars: list[Bar]) -> list[tuple[int, float]]:
+    projected: list[tuple[int, float]] = []
+    point_index = 0
+    current: float | None = None
+    for bar in bars:
+        while point_index < len(points) and points[point_index][0] <= bar.t:
+            current = points[point_index][1]
+            point_index += 1
+        if current is not None:
+            projected.append((bar.t, current))
+    return projected

app/api/routes.py

@@ -31,3 +31,15 @@ def bars(request: Request, tf: str = "1m", limit: int = Query(500, ge=1, le=5000
         raise HTTPException(400, "Unknown timeframe") from exc
     values = request.app.state.runtime.store.get(timeframe, limit)
     return {"tf": timeframe.value, "bars": [bar.to_dict() for bar in values]}
+
+
+@router.get("/levels")
+def levels(request: Request, tf: str = "all"):
+    values = request.app.state.runtime.levels
+    if tf != "all":
+        try:
+            timeframe = Timeframe(tf)
+        except ValueError as exc:
+            raise HTTPException(400, "Unknown timeframe") from exc
+        values = [level for level in values if level.tf is timeframe]
+    return {"levels": [level.to_dict() for level in values]}

app/api/ws.py

@@ -12,7 +12,7 @@ def snapshot(runtime, tf: Timeframe) -> dict:
         "type": "snapshot",
         "tf": tf.value,
         "bars": [bar.to_dict() for bar in runtime.store.get(tf, 1000)],
-        "levels": [],
+        "levels": [level.to_dict() for level in runtime.levels],
         "clusters": [],
         "price": runtime.store.get(Timeframe.M1, 1)[-1].c
         if runtime.store.get(Timeframe.M1, 1)
@@ -40,9 +40,15 @@ async def websocket_endpoint(websocket: WebSocket):
     receiver = asyncio.create_task(receive())
     try:
         while True:
-            bar = await queue.get()
+            event = await queue.get()
-            if bar.tf is tf:
+            if event["type"] == "bar" and event["bar"].tf is tf:
-                await websocket.send_json({"type": "bar", "tf": tf.value, "bar": bar.to_dict()})
+                await websocket.send_json(
+                    {"type": "bar", "tf": tf.value, "bar": event["bar"].to_dict()}
+                )
+            elif event["type"] == "levels":
+                await websocket.send_json(
+                    {"type": "levels", "levels": [level.to_dict() for level in event["levels"]]}
+                )
     except (WebSocketDisconnect, asyncio.CancelledError):
         pass
     finally:

app/config.py

@@ -45,3 +45,21 @@ class Settings(BaseSettings):
     @property
     def enabled_timeframes(self) -> list[Timeframe]:
         return [Timeframe(value.strip()) for value in self.timeframes.split(",") if value.strip()]
+
+    @property
+    def ma_sets(self) -> dict[Timeframe, list[tuple[str, int]]]:
+        configured = {
+            Timeframe.D1: self.ma_sets__1d,
+            Timeframe.H4: self.ma_sets__4h,
+            Timeframe.H1: self.ma_sets__1h,
+        }
+        result: dict[Timeframe, list[tuple[str, int]]] = {}
+        for tf, value in configured.items():
+            definitions = []
+            for item in filter(None, (part.strip().lower() for part in value.split(","))):
+                kind = "sma" if item.startswith("sma") else "ema" if item.startswith("ema") else ""
+                if not kind or not item[len(kind) :].isdigit():
+                    raise ValueError(f"Invalid MA definition: {item}")
+                definitions.append((kind, int(item[len(kind) :])))
+            result[tf] = definitions
+        return result

app/runtime.py

@@ -3,6 +3,8 @@ from dataclasses import dataclass, field
 
 from app.bars.models import Bar, Timeframe
 from app.bars.aggregator import Aggregator
+from app.analysis.levels import Level
+from app.analysis.moving_averages import build_ma_levels
 from app.bars.store import InMemoryBarStore
 from app.config import Settings
 from app.market.factory import live_source, seed_source
@@ -14,8 +16,9 @@ class Runtime:
     settings: Settings
     store: InMemoryBarStore = field(init=False)
     stream: StreamService = field(init=False)
-    subscribers: set[asyncio.Queue[Bar]] = field(default_factory=set)
+    subscribers: set[asyncio.Queue[dict]] = field(default_factory=set)
     aggregator: Aggregator = field(init=False)
+    levels: list[Level] = field(default_factory=list)
 
     def __post_init__(self) -> None:
         self.store = InMemoryBarStore(self.settings.max_bars_per_tf)
@@ -26,10 +29,22 @@ class Runtime:
     async def on_bar(self, bar: Bar) -> None:
         for aggregated in self.aggregator.update(bar):
             self.store.put(aggregated)
-            for queue in self.subscribers.copy():
+            self.broadcast({"type": "bar", "bar": aggregated})
-                if queue.full():
+            if self.settings.ma_sets.get(aggregated.tf):
-                    queue.get_nowait()
+                self.rebuild_levels()
-                queue.put_nowait(aggregated)
+
+    def broadcast(self, event: dict) -> None:
+        for queue in self.subscribers.copy():
+            if queue.full():
+                queue.get_nowait()
+            queue.put_nowait(event)
+
+    def rebuild_levels(self) -> None:
+        self.levels = build_ma_levels(
+            {tf: self.store.get(tf) for tf in self.settings.ma_sets},
+            self.settings.ma_sets,
+        )
+        self.broadcast({"type": "levels", "levels": self.levels})
 
     async def start(self) -> asyncio.Task:
         try:

static/app.js

@@ -30,11 +30,14 @@ createApp({
         const message = JSON.parse(data);
         if (message.type === 'snapshot') {
           chartApi.setBars(message.bars);
+          chartApi.syncLevels(message.levels || []);
           price.value = message.price;
         } else if (message.type === 'bar') {
           chartApi.updateBar(message.bar);
           price.value = message.bar.c;
           status.value.last_bar_t = message.bar.t;
+        } else if (message.type === 'levels') {
+          chartApi.syncLevels(message.levels);
         }
       };
       socket.onclose = () => {

static/chart.js

@@ -3,6 +3,7 @@ class ConfluenceChart {
     this.chart = null;
     this.candles = null;
     this.resizeObserver = null;
+    this.levelSeries = new Map();
   }
 
   create(el) {
@@ -36,6 +37,36 @@ class ConfluenceChart {
 
   updateBar(bar) { this.candles.update(this.toCandle(bar)); }
 
+  syncLevels(levels) {
+    const wanted = new Set(levels.filter(level => !level.hidden).map(level => level.id));
+    for (const [id, entry] of this.levelSeries) {
+      if (!wanted.has(id)) {
+        this.chart.removeSeries(entry.series);
+        this.levelSeries.delete(id);
+      }
+    }
+    for (const level of levels) {
+      if (level.hidden) continue;
+      let entry = this.levelSeries.get(level.id);
+      const options = {
+        color: ConfluenceChart.tfColors[level.tf],
+        lineWidth: level.tf === '1d' ? 2 : 1,
+        lineType: LightweightCharts.LineType.WithSteps,
+        lineStyle: level.provisional ? LightweightCharts.LineStyle.Dashed : LightweightCharts.LineStyle.Solid,
+        priceLineVisible: false,
+        lastValueVisible: true,
+        title: level.label,
+      };
+      if (!entry) {
+        entry = { series: this.chart.addSeries(LightweightCharts.LineSeries, options) };
+        this.levelSeries.set(level.id, entry);
+      } else {
+        entry.series.applyOptions(options);
+      }
+      entry.series.setData((level.points || []).map(([time, value]) => ({ time, value })));
+    }
+  }
+
   toCandle(bar) {
     return { time: bar.t, open: bar.o, high: bar.h, low: bar.l, close: bar.c };
   }
@@ -46,4 +77,9 @@ class ConfluenceChart {
   }
 }
 
+ConfluenceChart.tfColors = {
+  '1m':'#82909f', '2m':'#8a92df', '5m':'#65b7cf', '15m':'#45c39b',
+  '30m':'#a8c85d', '1h':'#efb643', '4h':'#ec7b42', '1d':'#d96073',
+};
+
 window.ConfluenceChart = ConfluenceChart;

tests/test_moving_averages.py

@@ -0,0 +1,61 @@
+from datetime import datetime, timedelta
+from zoneinfo import ZoneInfo
+
+from app.analysis.moving_averages import build_ma_levels, project_step
+from app.bars.models import Bar, Timeframe
+
+ET = ZoneInfo("America/New_York")
+
+
+def daily_bars(count: int, last_forming: bool = False) -> list[Bar]:
+    start = datetime(2025, 1, 5, 18, tzinfo=ET)
+    return [
+        Bar(
+            Timeframe.D1,
+            int((start + timedelta(days=index)).timestamp()),
+            index + 1,
+            index + 2,
+            index,
+            index + 1,
+            100,
+            not (last_forming and index == count - 1),
+            "ES=F",
+            "replay",
+        )
+        for index in range(count)
+    ]
+
+
+def test_daily_sma_set_has_known_values_and_stable_ids():
+    bars = daily_bars(210)
+    definitions = {Timeframe.D1: [("sma", p) for p in (10, 20, 50, 100, 200)]}
+    levels = build_ma_levels({Timeframe.D1: bars}, definitions)
+
+    assert [level.id for level in levels] == [f"ma:1d:sma:{p}" for p in (10, 20, 50, 100, 200)]
+    assert [level.anchor_p for level in levels] == [205.5, 200.5, 185.5, 160.5, 110.5]
+
+
+def test_nothing_emitted_before_closed_bar_warmup():
+    bars = daily_bars(200, last_forming=True)
+    assert build_ma_levels({Timeframe.D1: bars}, {Timeframe.D1: [("sma", 200)]}) == []
+
+
+def test_forming_daily_value_is_provisional():
+    bars = daily_bars(201, last_forming=True)
+    level = build_ma_levels({Timeframe.D1: bars}, {Timeframe.D1: [("sma", 200)]})[0]
+    assert level.provisional is True
+
+
+def test_step_projection_changes_only_at_session_boundary():
+    session_one = daily_bars(1)[0].t
+    session_two = daily_bars(2)[1].t
+    minutes = [
+        Bar(Timeframe.M1, t, 1, 1, 1, 1, 0, True, "ES=F", "replay")
+        for t in (session_one, session_one + 60, session_two - 60, session_two)
+    ]
+    assert project_step([(session_one, 100), (session_two, 101)], minutes) == [
+        (session_one, 100),
+        (session_one + 60, 100),
+        (session_two - 60, 100),
+        (session_two, 101),
+    ]