jcback 0.1.1__tar.gz

jcback-0.1.1/PKG-INFO

@@ -0,0 +1,168 @@
+Metadata-Version: 2.4
+Name: jcback
+Version: 0.1.1
+Summary: Event-driven bar replay backtest engine for China A-share, ETF, bond and futures
+Author: jcback contributors
+License-Expression: MIT
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+Requires-Dist: loguru>=0.7
+Requires-Dist: numpy>=1.24
+Requires-Dist: pandas>=2.0
+Requires-Dist: PyYAML>=6.0
+Provides-Extra: jcdata
+Requires-Dist: jcdata>=0.1; extra == "jcdata"
+Provides-Extra: dev
+Requires-Dist: pytest>=7; extra == "dev"
+
+# jcback
+
+事件驱动、逐 K 线推进的 A 股 / ETF / 可转债 / 期货回测库。
+
+支持函数式策略（`init` / `on_bar`，与**几策**平台写法一致）与类式策略（`Strategy` 子类），可自带 OHLCV、使用内置模拟数据，或通过 **[jcdata](https://pypi.org/project/jcdata)** 拉取行情。
+
+## 安装
+
+```bash
+pip install jcback
+```
+
+需要 jcdata 行情时：
+
+```bash
+pip install jcback[jcdata]
+```
+
+或单独安装：
+
+```bash
+pip install jcdata
+```
+
+要求 **Python >= 3.10**。
+
+## 快速开始
+
+### 函数式策略 + 模拟数据
+
+无需外部行情，适合本地验证逻辑：
+
+```python
+from jcback import run, MODE_BACKTEST, MATCH_NEXT_OPEN, ADJUST_PREV
+
+
+def init(context):
+    context.subscribe(context.symbols, frequency="1d", wait_group=True)
+
+
+def on_bar(context, bars):
+    for bar in bars:
+        context.order_target_percent(bar.symbol, 0.5)
+
+
+result = run(
+    strategy_id="demo",
+    mode=MODE_BACKTEST,
+    symbols="600000.SH",
+    backtest_start_time="2025-01-02",
+    backtest_end_time="2025-03-31",
+    backtest_adjust=ADJUST_PREV,
+    backtest_match_mode=MATCH_NEXT_OPEN,
+    init_func=init,
+    on_bar=on_bar,
+    use_mock=True,
+)
+
+print(result.indicator["return_pct"], "%")
+```
+
+### 对接 jcdata 实盘行情
+
+```python
+import jcdata
+from jcback import run, MODE_BACKTEST, MATCH_NEXT_OPEN, print_backtest_report, setup_logger
+
+jcdata.login()  # 或 jcdata.login(token="YOUR_TOKEN", base_url="...")
+
+setup_logger()
+
+result = run(
+    strategy_id="demo",
+    mode=MODE_BACKTEST,
+    symbols=["600000.SH", "601318.SH"],
+    backtest_start_time="2025-01-02",
+    backtest_end_time="2025-06-30",
+    backtest_match_mode=MATCH_NEXT_OPEN,
+    init_func=init,
+    on_bar=on_bar,
+    use_jcdata=True,
+)
+
+print_backtest_report(result, "demo")
+```
+
+### 类式策略
+
+```python
+from jcback import Strategy, run_backtest
+
+
+class SmaCross(Strategy):
+    n_short, n_long = 5, 20
+
+    def init(self):
+        c = self.data.Close
+        self.ma_s = self.I(lambda x, n: x.rolling(n).mean(), c, self.n_short)
+        self.ma_l = self.I(lambda x, n: x.rolling(n).mean(), c, self.n_long)
+
+    def next(self):
+        if self.ma_s[-2] < self.ma_l[-2] and self.ma_s[-1] > self.ma_l[-1]:
+            self.buy()
+        elif self.ma_s[-2] > self.ma_l[-2] and self.ma_s[-1] < self.ma_l[-1]:
+            self.position.close()
+
+
+result = run_backtest(
+    SmaCross,
+    symbols="600000.SH",
+    start_date="2025-01-02",
+    end_date="2025-06-30",
+    use_mock=True,
+)
+```
+
+### 自备 OHLCV
+
+`data` 列为 `Open` / `High` / `Low` / `Close` / `Volume`（可选 `Amount`），索引为时间：
+
+```python
+run(..., data=ohlcv_df, symbols="600000.SH")
+# 多标的：data={"600000.SH": df1, "601318.SH": df2}
+```
+
+## 常用常量
+
+| 常量 | 含义 |
+|------|------|
+| `MODE_BACKTEST` | 回测模式 |
+| `ADJUST_NONE` / `ADJUST_PREV` / `ADJUST_POST` | 不复权 / 前复权 / 后复权 |
+| `MATCH_NEXT_OPEN` | 下一根 K 线开盘价撮合 |
+| `MATCH_CURRENT_CLOSE` | 当前 K 线收盘价撮合 |
+
+## 手续费与交割
+
+未显式传入 `commission` 时，按包内 `config.yaml` 及 `product_fees.yaml` 的品类规则计算（股票、ETF、转债、期货等）。
+
+`settlement_rule="T+1"`（股票默认）或 `"T+0"`（转债、期货等）控制当日买入是否可卖。
+
+## 主要导出
+
+`run`、`run_backtest`、`Strategy`、`Backtest`、`Context`、`Bar`、`resolve_config`、  
+`load_from_jcdata`、`load_multi_ohlcv`、`jcdata_to_ohlcv`、`make_mock_jcdata`、  
+`get_last_result`、`serialize_for_ui`、`print_backtest_report`、`setup_logger`
+
+完整 API 见 `jcback.__all__` 或接口文档。
+
+## 许可证
+
+MIT

jcback-0.1.1/README.md

@@ -0,0 +1,151 @@
+# jcback
+
+事件驱动、逐 K 线推进的 A 股 / ETF / 可转债 / 期货回测库。
+
+支持函数式策略（`init` / `on_bar`，与**几策**平台写法一致）与类式策略（`Strategy` 子类），可自带 OHLCV、使用内置模拟数据，或通过 **[jcdata](https://pypi.org/project/jcdata)** 拉取行情。
+
+## 安装
+
+```bash
+pip install jcback
+```
+
+需要 jcdata 行情时：
+
+```bash
+pip install jcback[jcdata]
+```
+
+或单独安装：
+
+```bash
+pip install jcdata
+```
+
+要求 **Python >= 3.10**。
+
+## 快速开始
+
+### 函数式策略 + 模拟数据
+
+无需外部行情，适合本地验证逻辑：
+
+```python
+from jcback import run, MODE_BACKTEST, MATCH_NEXT_OPEN, ADJUST_PREV
+
+
+def init(context):
+    context.subscribe(context.symbols, frequency="1d", wait_group=True)
+
+
+def on_bar(context, bars):
+    for bar in bars:
+        context.order_target_percent(bar.symbol, 0.5)
+
+
+result = run(
+    strategy_id="demo",
+    mode=MODE_BACKTEST,
+    symbols="600000.SH",
+    backtest_start_time="2025-01-02",
+    backtest_end_time="2025-03-31",
+    backtest_adjust=ADJUST_PREV,
+    backtest_match_mode=MATCH_NEXT_OPEN,
+    init_func=init,
+    on_bar=on_bar,
+    use_mock=True,
+)
+
+print(result.indicator["return_pct"], "%")
+```
+
+### 对接 jcdata 实盘行情
+
+```python
+import jcdata
+from jcback import run, MODE_BACKTEST, MATCH_NEXT_OPEN, print_backtest_report, setup_logger
+
+jcdata.login()  # 或 jcdata.login(token="YOUR_TOKEN", base_url="...")
+
+setup_logger()
+
+result = run(
+    strategy_id="demo",
+    mode=MODE_BACKTEST,
+    symbols=["600000.SH", "601318.SH"],
+    backtest_start_time="2025-01-02",
+    backtest_end_time="2025-06-30",
+    backtest_match_mode=MATCH_NEXT_OPEN,
+    init_func=init,
+    on_bar=on_bar,
+    use_jcdata=True,
+)
+
+print_backtest_report(result, "demo")
+```
+
+### 类式策略
+
+```python
+from jcback import Strategy, run_backtest
+
+
+class SmaCross(Strategy):
+    n_short, n_long = 5, 20
+
+    def init(self):
+        c = self.data.Close
+        self.ma_s = self.I(lambda x, n: x.rolling(n).mean(), c, self.n_short)
+        self.ma_l = self.I(lambda x, n: x.rolling(n).mean(), c, self.n_long)
+
+    def next(self):
+        if self.ma_s[-2] < self.ma_l[-2] and self.ma_s[-1] > self.ma_l[-1]:
+            self.buy()
+        elif self.ma_s[-2] > self.ma_l[-2] and self.ma_s[-1] < self.ma_l[-1]:
+            self.position.close()
+
+
+result = run_backtest(
+    SmaCross,
+    symbols="600000.SH",
+    start_date="2025-01-02",
+    end_date="2025-06-30",
+    use_mock=True,
+)
+```
+
+### 自备 OHLCV
+
+`data` 列为 `Open` / `High` / `Low` / `Close` / `Volume`（可选 `Amount`），索引为时间：
+
+```python
+run(..., data=ohlcv_df, symbols="600000.SH")
+# 多标的：data={"600000.SH": df1, "601318.SH": df2}
+```
+
+## 常用常量
+
+| 常量 | 含义 |
+|------|------|
+| `MODE_BACKTEST` | 回测模式 |
+| `ADJUST_NONE` / `ADJUST_PREV` / `ADJUST_POST` | 不复权 / 前复权 / 后复权 |
+| `MATCH_NEXT_OPEN` | 下一根 K 线开盘价撮合 |
+| `MATCH_CURRENT_CLOSE` | 当前 K 线收盘价撮合 |
+
+## 手续费与交割
+
+未显式传入 `commission` 时，按包内 `config.yaml` 及 `product_fees.yaml` 的品类规则计算（股票、ETF、转债、期货等）。
+
+`settlement_rule="T+1"`（股票默认）或 `"T+0"`（转债、期货等）控制当日买入是否可卖。
+
+## 主要导出
+
+`run`、`run_backtest`、`Strategy`、`Backtest`、`Context`、`Bar`、`resolve_config`、  
+`load_from_jcdata`、`load_multi_ohlcv`、`jcdata_to_ohlcv`、`make_mock_jcdata`、  
+`get_last_result`、`serialize_for_ui`、`print_backtest_report`、`setup_logger`
+
+完整 API 见 `jcback.__all__` 或接口文档。
+
+## 许可证
+
+MIT

jcback-0.1.1/__init__.py

@@ -0,0 +1,89 @@
+"""
+jcback — 事件驱动 A 股回测框架。
+
+函数式策略（推荐，与几策平台 init / on_bar 写法一致）::
+
+    def init(context): ...
+    def on_bar(context, bars): ...
+
+    def run_main():
+        run(
+            strategy_id="demo",
+            mode=MODE_BACKTEST,
+            symbols="600000.SH",
+            backtest_start_time="2025-01-02",
+            backtest_end_time="2025-03-31",
+            backtest_match_mode=MATCH_NEXT_OPEN,
+            init_func=init,
+            on_bar=on_bar,
+            use_mock=True,
+        )
+
+类式策略（jcback.Strategy 子类）::
+
+    from jcback import run_backtest, Strategy
+    result = run_backtest(MyStrategy, use_mock=True, symbols="600000.SH", start_date="2025-01-02", ...)
+"""
+from __future__ import annotations
+
+from .sim.facade import Backtest, Strategy
+from .runtime import (
+    ADJUST_NONE,
+    ADJUST_POST,
+    ADJUST_PREV,
+    Bar,
+    BacktestConfig,
+    BacktestEvent,
+    BacktestResult,
+    Context,
+    EventBacktest,
+    EventBus,
+    FunctionStrategy,
+    MATCH_CURRENT_CLOSE,
+    MATCH_NEXT_OPEN,
+    MODE_BACKTEST,
+    infer_asset_type,
+    jcdata_to_ohlcv,
+    load_from_jcdata,
+    load_multi_ohlcv,
+    make_mock_jcdata,
+    parse_symbols,
+    resolve_config,
+)
+from .runtime.export import emit_report_marker, get_last_result, serialize_for_ui
+from .runtime.report import print_backtest_report, setup_logger
+from .entry import default_backtest_window, run, run_backtest
+
+__all__ = [
+    "ADJUST_NONE",
+    "ADJUST_POST",
+    "ADJUST_PREV",
+    "Backtest",
+    "BacktestConfig",
+    "BacktestEvent",
+    "BacktestResult",
+    "Bar",
+    "Context",
+    "EventBacktest",
+    "EventBus",
+    "FunctionStrategy",
+    "MATCH_CURRENT_CLOSE",
+    "MATCH_NEXT_OPEN",
+    "MODE_BACKTEST",
+    "Strategy",
+    "default_backtest_window",
+    "infer_asset_type",
+    "jcdata_to_ohlcv",
+    "load_from_jcdata",
+    "load_multi_ohlcv",
+    "make_mock_jcdata",
+    "parse_symbols",
+    "emit_report_marker",
+    "get_last_result",
+    "print_backtest_report",
+    "resolve_config",
+    "run",
+    "run_backtest",
+    "serialize_for_ui",
+    "setup_logger",
+]

jcback-0.1.1/config.yaml

@@ -0,0 +1,56 @@
+# jcback 回测默认配置
+# 手续费规则：按品种区分买卖、期货平昨/平今；支持固定费 + 成交额比例
+
+commission:
+  # 未命中品种规则时的兜底
+  default:
+    buy_ratio: 0.0001
+    sell_ratio: 0.0001
+    fixed: 0.0
+    min_commission: 0.0
+    stamp_tax_ratio: 0.0
+
+  by_asset:
+    # A 股：买卖佣金不同，卖出另收印花税
+    stock:
+      buy_ratio: 0.00025
+      sell_ratio: 0.00025
+      fixed: 0.0
+      min_commission: 5.0
+      stamp_tax_ratio: 0.001
+
+    etf:
+      buy_ratio: 0.00025
+      sell_ratio: 0.00025
+      fixed: 0.0
+      min_commission: 5.0
+      stamp_tax_ratio: 0.0
+
+    # 可转债：买卖费率通常一致，无印花税
+    convertible_bond:
+      buy_ratio: 0.00005
+      sell_ratio: 0.00005
+      fixed: 0.0
+      min_commission: 0.0
+      stamp_tax_ratio: 0.0
+
+    bond:
+      buy_ratio: 0.00005
+      sell_ratio: 0.00005
+      fixed: 0.0
+      min_commission: 0.0
+      stamp_tax_ratio: 0.0
+
+    # 期货：通用兜底（未在 product_fees.yaml 列出的品种使用）
+    future:
+      multiplier: 1.0
+      open_ratio: 0.000023
+      open_fixed: 0.0
+      close_yesterday_ratio: 0.000023
+      close_yesterday_fixed: 0.0
+      close_today_ratio: 0.00023
+      close_today_fixed: 0.0
+      min_commission: 0.0
+
+  # 中国期货各品种交易所标准费率表（按品种代码 IF/RB/MA 等匹配）
+  future_fee_table: product_fees.yaml