fce4d536ea
Added HOLD candles as negative samples to increase training data from ~535 to ~3,200 samples. Introduced a negative_ratio parameter in generate_dataset_vectorized() for sampling HOLD candles alongside signal candles. Implemented stratified undersampling to ensure signal samples are preserved during training. Updated relevant tests to validate new functionality and maintain compatibility with existing tests. - Modified dataset_builder.py to include HOLD negative sampling logic - Updated train_model.py to apply stratified undersampling - Added tests for new sampling methods Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
249 lines
9.6 KiB
Python
249 lines
9.6 KiB
Python
import pytest
|
|
from unittest.mock import AsyncMock, MagicMock, patch
|
|
import pandas as pd
|
|
import numpy as np
|
|
import os
|
|
from src.bot import TradingBot
|
|
from src.config import Config
|
|
|
|
|
|
@pytest.fixture
|
|
def config():
|
|
os.environ.update({
|
|
"BINANCE_API_KEY": "k",
|
|
"BINANCE_API_SECRET": "s",
|
|
"SYMBOL": "XRPUSDT",
|
|
"LEVERAGE": "10",
|
|
"RISK_PER_TRADE": "0.02",
|
|
"NOTION_TOKEN": "secret_test",
|
|
"NOTION_DATABASE_ID": "db_test",
|
|
"DISCORD_WEBHOOK_URL": "",
|
|
})
|
|
return Config()
|
|
|
|
|
|
@pytest.fixture
|
|
def sample_df():
|
|
np.random.seed(0)
|
|
n = 100
|
|
close = np.cumsum(np.random.randn(n) * 0.01) + 0.5
|
|
return pd.DataFrame({
|
|
"open": close,
|
|
"high": close * 1.005,
|
|
"low": close * 0.995,
|
|
"close": close,
|
|
"volume": np.random.randint(100000, 1000000, n).astype(float),
|
|
})
|
|
|
|
|
|
def test_bot_uses_multi_symbol_stream(config):
|
|
from src.data_stream import MultiSymbolStream
|
|
with patch("src.bot.BinanceFuturesClient"):
|
|
bot = TradingBot(config)
|
|
assert isinstance(bot.stream, MultiSymbolStream)
|
|
|
|
def test_bot_stream_has_btc_eth_buffers(config):
|
|
with patch("src.bot.BinanceFuturesClient"):
|
|
bot = TradingBot(config)
|
|
assert "btcusdt" in bot.stream.buffers
|
|
assert "ethusdt" in bot.stream.buffers
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_bot_processes_signal(config, sample_df):
|
|
with patch("src.bot.BinanceFuturesClient") as MockExchange:
|
|
MockExchange.return_value = AsyncMock()
|
|
bot = TradingBot(config)
|
|
|
|
bot.exchange = AsyncMock()
|
|
bot.exchange.get_balance = AsyncMock(return_value=1000.0)
|
|
bot.exchange.get_position = AsyncMock(return_value=None)
|
|
bot.exchange.place_order = AsyncMock(return_value={"orderId": "123"})
|
|
bot.exchange.set_leverage = AsyncMock(return_value={})
|
|
bot.exchange.calculate_quantity = MagicMock(return_value=100.0)
|
|
bot.exchange.MIN_NOTIONAL = 5.0
|
|
|
|
with patch("src.bot.Indicators") as MockInd:
|
|
mock_ind = MagicMock()
|
|
mock_ind.calculate_all.return_value = sample_df
|
|
mock_ind.get_signal.return_value = "LONG"
|
|
mock_ind.get_atr_stop.return_value = (0.48, 0.56)
|
|
MockInd.return_value = mock_ind
|
|
await bot.process_candle(sample_df)
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_close_and_reenter_calls_open_when_ml_passes(config, sample_df):
|
|
"""반대 시그널 + ML 필터 통과 시 청산 후 재진입해야 한다."""
|
|
with patch("src.bot.BinanceFuturesClient"):
|
|
bot = TradingBot(config)
|
|
|
|
bot._close_position = AsyncMock()
|
|
bot._open_position = AsyncMock()
|
|
bot.risk = MagicMock()
|
|
bot.risk.can_open_new_position.return_value = True
|
|
bot.ml_filter = MagicMock()
|
|
bot.ml_filter.is_model_loaded.return_value = True
|
|
bot.ml_filter.should_enter.return_value = True
|
|
|
|
position = {"positionAmt": "100", "entryPrice": "0.5", "markPrice": "0.52"}
|
|
await bot._close_and_reenter(position, "SHORT", sample_df)
|
|
|
|
bot._close_position.assert_awaited_once_with(position)
|
|
bot._open_position.assert_awaited_once_with("SHORT", sample_df)
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_close_and_reenter_skips_open_when_ml_blocks(config, sample_df):
|
|
"""ML 필터 차단 시 청산만 하고 재진입하지 않아야 한다."""
|
|
with patch("src.bot.BinanceFuturesClient"):
|
|
bot = TradingBot(config)
|
|
|
|
bot._close_position = AsyncMock()
|
|
bot._open_position = AsyncMock()
|
|
bot.ml_filter = MagicMock()
|
|
bot.ml_filter.is_model_loaded.return_value = True
|
|
bot.ml_filter.should_enter.return_value = False
|
|
|
|
position = {"positionAmt": "100", "entryPrice": "0.5", "markPrice": "0.52"}
|
|
await bot._close_and_reenter(position, "SHORT", sample_df)
|
|
|
|
bot._close_position.assert_awaited_once_with(position)
|
|
bot._open_position.assert_not_called()
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_close_and_reenter_skips_open_when_max_positions_reached(config, sample_df):
|
|
"""최대 포지션 수 도달 시 청산만 하고 재진입하지 않아야 한다."""
|
|
with patch("src.bot.BinanceFuturesClient"):
|
|
bot = TradingBot(config)
|
|
|
|
bot._close_position = AsyncMock()
|
|
bot._open_position = AsyncMock()
|
|
bot.risk = MagicMock()
|
|
bot.risk.can_open_new_position.return_value = False
|
|
|
|
position = {"positionAmt": "100", "entryPrice": "0.5", "markPrice": "0.52"}
|
|
await bot._close_and_reenter(position, "SHORT", sample_df)
|
|
|
|
bot._close_position.assert_awaited_once_with(position)
|
|
bot._open_position.assert_not_called()
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_process_candle_calls_close_and_reenter_on_reverse_signal(config, sample_df):
|
|
"""반대 시그널 시 process_candle이 _close_and_reenter를 호출해야 한다."""
|
|
with patch("src.bot.BinanceFuturesClient"):
|
|
bot = TradingBot(config)
|
|
|
|
bot.exchange = AsyncMock()
|
|
bot.exchange.get_position = AsyncMock(return_value={
|
|
"positionAmt": "100",
|
|
"entryPrice": "0.5",
|
|
"markPrice": "0.52",
|
|
})
|
|
bot._close_and_reenter = AsyncMock()
|
|
bot.ml_filter = MagicMock()
|
|
bot.ml_filter.is_model_loaded.return_value = False
|
|
bot.ml_filter.should_enter.return_value = True
|
|
|
|
with patch("src.bot.Indicators") as MockInd:
|
|
mock_ind = MagicMock()
|
|
mock_ind.calculate_all.return_value = sample_df
|
|
mock_ind.get_signal.return_value = "SHORT" # 현재 LONG 포지션에 반대 시그널
|
|
MockInd.return_value = mock_ind
|
|
await bot.process_candle(sample_df)
|
|
|
|
bot._close_and_reenter.assert_awaited_once()
|
|
call_args = bot._close_and_reenter.call_args
|
|
assert call_args.args[1] == "SHORT"
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_process_candle_passes_raw_signal_to_close_and_reenter_even_if_ml_loaded(config, sample_df):
|
|
"""포지션 보유 시 ML 필터가 로드되어 있어도 process_candle은 raw signal을 _close_and_reenter에 전달한다."""
|
|
with patch("src.bot.BinanceFuturesClient"):
|
|
bot = TradingBot(config)
|
|
|
|
bot.exchange = AsyncMock()
|
|
bot.exchange.get_position = AsyncMock(return_value={
|
|
"positionAmt": "100",
|
|
"entryPrice": "0.5",
|
|
"markPrice": "0.52",
|
|
})
|
|
bot._close_and_reenter = AsyncMock()
|
|
bot.ml_filter = MagicMock()
|
|
bot.ml_filter.is_model_loaded.return_value = True # 모델 로드됨
|
|
bot.ml_filter.should_enter.return_value = False # ML이 차단하더라도
|
|
|
|
with patch("src.bot.Indicators") as MockInd:
|
|
mock_ind = MagicMock()
|
|
mock_ind.calculate_all.return_value = sample_df
|
|
mock_ind.get_signal.return_value = "SHORT"
|
|
MockInd.return_value = mock_ind
|
|
await bot.process_candle(sample_df)
|
|
|
|
# ML 필터가 차단해도 _close_and_reenter는 호출되어야 한다 (ML 재평가는 내부에서)
|
|
bot._close_and_reenter.assert_awaited_once()
|
|
call_args = bot._close_and_reenter.call_args
|
|
assert call_args.args[1] == "SHORT"
|
|
# process_candle에서 ml_filter.should_enter가 호출되지 않아야 한다
|
|
bot.ml_filter.should_enter.assert_not_called()
|
|
|
|
|
|
@pytest.mark.asyncio
|
|
async def test_process_candle_fetches_oi_and_funding(config, sample_df):
|
|
"""process_candle()이 OI와 펀딩비를 조회하고 build_features에 전달하는지 확인."""
|
|
with patch("src.bot.BinanceFuturesClient"):
|
|
bot = TradingBot(config)
|
|
|
|
bot.exchange = AsyncMock()
|
|
bot.exchange.get_balance = AsyncMock(return_value=1000.0)
|
|
bot.exchange.get_position = AsyncMock(return_value=None)
|
|
bot.exchange.place_order = AsyncMock(return_value={"orderId": "1"})
|
|
bot.exchange.set_leverage = AsyncMock()
|
|
bot.exchange.get_open_interest = AsyncMock(return_value=5000000.0)
|
|
bot.exchange.get_funding_rate = AsyncMock(return_value=0.0001)
|
|
|
|
# 신호를 LONG으로 강제해 build_features가 반드시 호출되도록 함
|
|
with patch("src.bot.Indicators") as mock_ind_cls:
|
|
mock_ind = MagicMock()
|
|
mock_ind.calculate_all.return_value = sample_df
|
|
mock_ind.get_signal.return_value = "LONG"
|
|
mock_ind_cls.return_value = mock_ind
|
|
|
|
with patch("src.bot.build_features") as mock_build:
|
|
from src.ml_features import FEATURE_COLS
|
|
mock_build.return_value = pd.Series({col: 0.0 for col in FEATURE_COLS})
|
|
bot.ml_filter.is_model_loaded = MagicMock(return_value=False)
|
|
# _open_position은 이 테스트의 관심사가 아니므로 mock 처리
|
|
bot._open_position = AsyncMock()
|
|
await bot.process_candle(sample_df)
|
|
|
|
assert mock_build.called
|
|
call_kwargs = mock_build.call_args.kwargs
|
|
assert "oi_change" in call_kwargs
|
|
assert "funding_rate" in call_kwargs
|
|
|
|
|
|
def test_calc_oi_change_first_candle_returns_zero(config):
|
|
"""첫 캔들은 0.0을 반환하고 _prev_oi를 설정한다."""
|
|
with patch("src.bot.BinanceFuturesClient"):
|
|
bot = TradingBot(config)
|
|
assert bot._calc_oi_change(5000000.0) == 0.0
|
|
assert bot._prev_oi == 5000000.0
|
|
|
|
|
|
def test_calc_oi_change_api_failure_does_not_corrupt_state(config):
|
|
"""API 실패 시 _fetch_market_microstructure가 _calc_oi_change를 호출하지 않아 상태가 오염되지 않는다."""
|
|
with patch("src.bot.BinanceFuturesClient"):
|
|
bot = TradingBot(config)
|
|
bot._prev_oi = 5000000.0
|
|
# API 실패 시 _fetch_market_microstructure는 oi_val > 0 체크로 _calc_oi_change를 건너뜀
|
|
# _calc_oi_change(0.0)을 직접 호출하면 _prev_oi가 0.0으로 오염되는 이전 버그를 재현
|
|
# 수정 후에는 _fetch_market_microstructure에서 0.0을 직접 반환하므로 이 경로가 없음
|
|
# 대신 _calc_oi_change가 정상 값에서만 호출되는지 확인
|
|
result = bot._calc_oi_change(5100000.0)
|
|
assert abs(result - 0.02) < 1e-6 # (5100000 - 5000000) / 5000000 = 0.02
|
|
assert bot._prev_oi == 5100000.0
|