| |
|
| | from __future__ import annotations |
| | from typing import Optional, Literal |
| | import pandas as pd |
| | from datetime import datetime |
| | from .models import TradePlan, PositionState |
| |
|
| | def _hits_level(low: float, high: float, level: float) -> bool: |
| | return low <= level <= high |
| |
|
| | def _decide_exit_on_bar( |
| | side: Literal["long", "short"], low: float, high: float, target: float, stoploss: float, |
| | tiebreaker: Literal["stoploss_first", "target_first"] = "stoploss_first", |
| | ) -> Optional[dict]: |
| | if side == "long": |
| | hit_target = high >= target |
| | hit_stop = low <= stoploss |
| | else: |
| | hit_target = low <= target |
| | hit_stop = high >= stoploss |
| |
|
| | if not hit_target and not hit_stop: |
| | return None |
| | if hit_target and hit_stop: |
| | return {"reason": "stoploss" if tiebreaker == "stoploss_first" else "target", |
| | "price": stoploss if tiebreaker == "stoploss_first" else target} |
| | return {"reason": "target", "price": target} if hit_target else {"reason": "stoploss", "price": stoploss} |
| |
|
| | def simulate_trade_from_signal( |
| | df: pd.DataFrame, |
| | trade: TradePlan, |
| | *, |
| | dt_col: str = "datetime", |
| | tiebreaker: Literal["stoploss_first", "target_first"] = "stoploss_first", |
| | lookback_minutes: int = 60, |
| | state: PositionState, |
| | ) -> PositionState: |
| | |
| | required = {"open","high","low","close", dt_col} |
| | if missing := [c for c in required if c not in df.columns]: |
| | raise ValueError(f"Missing columns: {missing}") |
| |
|
| | if trade.status == "No trade": |
| | return state |
| |
|
| | side = trade.type |
| | entry_at = float(trade.entry_at) |
| | target = float(trade.target) |
| | stoploss = float(trade.stoploss) |
| |
|
| | ts = pd.to_datetime(df[dt_col]) |
| |
|
| | entered = state.entered |
| | exited = state.exited |
| | entry_time = state.entry_time |
| | entry_price = state.entry_price |
| | exit_time = state.exit_time |
| | exit_price = state.exit_price |
| | exit_reason = state.exit_reason |
| |
|
| | if not entered: |
| | if side=="long" and entry_at<=stoploss: |
| | return PositionState( |
| | entered=entered, entry_time=entry_time, entry_price=entry_price, side=side, |
| | exited=exited, exit_time=exit_time, exit_price=exit_price, exit_reason=exit_reason, |
| | pnl_pct=state.pnl_pct, open_position=state.open_position, unrealized_pct=state.unrealized_pct, |
| | note="Error in pricing: Stoploss is higher than entry price for long side position, please adjust stoploss or entry price" |
| | ) |
| | elif side=="short" and entry_at>=stoploss: |
| | return PositionState( |
| | entered=entered, entry_time=entry_time, entry_price=entry_price, side=side, |
| | exited=exited, exit_time=exit_time, exit_price=exit_price, exit_reason=exit_reason, |
| | pnl_pct=state.pnl_pct, open_position=state.open_position, unrealized_pct=state.unrealized_pct, |
| | note="Error in pricing: Stoploss is lower than entry price for short side position, please adjust stoploss or entry price" |
| | ) |
| |
|
| | |
| | start_idx = max(0, abs(len(df) - int(lookback_minutes))) |
| |
|
| | for i in range(start_idx, len(df)): |
| | row = df.iloc[i] |
| | time_i, o, h, l, c = ts.iloc[i], float(row['open']), float(row['high']), float(row['low']), float(row['close']) |
| |
|
| | if not entered: |
| | if _hits_level(l, h, entry_at): |
| | entered = True |
| | entry_time = time_i |
| | entry_price = entry_at |
| | out = _decide_exit_on_bar(side, l, h, target, stoploss, tiebreaker) |
| | if out: |
| | exited = True |
| | exit_time = time_i |
| | exit_price = out["price"] |
| | exit_reason = out["reason"] |
| | break |
| | else: |
| | out = _decide_exit_on_bar(side, l, h, target, stoploss, tiebreaker) |
| | if out: |
| | exited = True |
| | exit_time = time_i |
| | exit_price = out["price"] |
| | exit_reason = out["reason"] |
| | break |
| |
|
| | def _pnl_pct(entry: float, exit_: float, side_: str) -> float: |
| | raw = (exit_ - entry) / entry |
| | return (raw if side_ == "long" else -raw) * 100.0 |
| |
|
| | if entered and exited: |
| | pnl = round(_pnl_pct(entry_price, exit_price, side), 4) |
| | return PositionState( |
| | entered=True, entry_time=entry_time, entry_price=entry_price, side=side, |
| | exited=True, exit_time=exit_time, exit_price=exit_price, exit_reason=exit_reason, |
| | pnl_pct=pnl, open_position=False, unrealized_pct=None, |
| | ) |
| |
|
| | if entered and not exited: |
| | last_close = float(df.iloc[-1]["close"]) if len(df) else entry_price |
| | upnl = round(_pnl_pct(entry_price, last_close, side), 4) |
| | return PositionState( |
| | entered=True, entry_time=entry_time, entry_price=entry_price, side=side, |
| | exited=False, open_position=True, unrealized_pct=upnl, |
| | ) |
| |
|
| | return state |
| |
|
| | def slice_intraday(df_1m: pd.DataFrame, start: datetime, end: datetime, dt_col: str = "datetime") -> pd.DataFrame: |
| | mask = (df_1m[dt_col] >= start) & (df_1m[dt_col] < end) |
| | out = df_1m.loc[mask].copy() |
| | out.reset_index(drop=True, inplace=True) |
| | return out |
| |
|