Skip to main content

A high-performance vector backtesting framework for quantitative strategies

Project description

Kepler Echo

Python License

向量化回测框架。

安装

pip install kepler-echo

快速开始

import pandas as pd
from kepler.echo import Strategy

# 价格数据 (MultiIndex: date, item)
price_data = []
for date in ['2020-01-01', '2020-01-02']:
    for stock, o, c in [('A', 10, 10.5), ('B', 20, 20.5), ('C', 30, 30.5)]:
        price_data.append({'date': date, 'item': stock, 'open': o, 'close': c})
price = pd.DataFrame(price_data).set_index(['date', 'item'])

# 信号
signal = pd.DataFrame({
    'A': [0.5, 0.6],
    'B': [-0.3, -0.2],
}, index=pd.date_range('2020-01-01', periods=2))

# 回测
result = (
    Strategy(begin="2020-01-01", end="2020-12-31")
    .data(price)
    .signal(signal)
    .commission((0.001, 0.001))
    .run()
)

print(result.nav)

API

Strategy

Strategy(
    begin="2001-01-01",        # 开始日期
    end="今天",                 # 结束日期
    matching="next_bar",       # 撮合: next_bar / current_bar
    benchmark="",              # 基准 (数据中的某列)
    commission=(0, 0),         # 手续费 (做多, 做空)
)

方法

方法 说明
.data(df, exec_price='open') 添加价格数据
.signal(df) 添加信号
.commission((long, short)) 设置手续费
.benchmark(symbol) 设置基准
.run() 运行,返回结果
.plot(log=True) 绘图

数据格式

支持两种格式:

1. pandas DataFrame (MultiIndex)

index 为 ['date', 'item'],columns 必须包含 closeexec_price 指定的列:

                      close  open
date       item
2020-01-01 A        10.5    10
           B        20.5    20
2020-01-02 A        11.0    10.5
           B        21.0    20.5

2. xarray DataArray

三维数组,维度为 (date, item, feature)

import xarray as xr
import numpy as np

# 创建 xarray DataArray
dates = pd.date_range('2020-01-01', periods=2)
items = ['A', 'B']
features = ['open', 'close']

data = xr.DataArray(
    np.random.randn(2, 2, 2),
    dims=['date', 'item', 'feature'],
    coords={'date': dates, 'item': items, 'feature': features}
)

# 使用
result = Strategy().data(data, exec_price='open').signal(signal).run()

信号格式

宽格式:

signal = pd.DataFrame({
    '000001.SZ': [0.5, 0.6],
    '000002.SZ': [-0.3, -0.2],
}, index=pd.date_range('2020-01-01', periods=2))

长格式:

signal = pd.DataFrame({
    'date': ['2020-01-01', '2020-01-01'],
    'stockid': ['000001.SZ', '000002.SZ'],
    'weight': [0.5, -0.3]
})

结果

result.nav      # 净值 DataFrame
result.hold     # 最终持仓
result.signal   # 原始信号
result.stats    # 统计 (turnover)

nav 列说明:

列名 说明
strategy 策略净值
{benchmark} 基准净值(如果设置了 benchmark)
relative 相对净值 = strategy / benchmark(如果设置了 benchmark)
drawdown 动态回撤(相对收益的回撤,或绝对收益的回撤)

撮合方式

  • next_bar: 下一根 K 线的 exec_price 价格(默认)
  • current_bar: 当前 K 线收盘价

执行价格

exec_price 参数指定 next_bar 模式下的执行价格列:

.data(price)                       # 使用开盘价(默认)
.data(price, exec_price='vwap')    # 使用 VWAP
.data(price, exec_price='close')   # 使用收盘价

许可证

GPL-3.0-or-later

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

kepler_echo-0.3.0.tar.gz (28.2 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

kepler_echo-0.3.0-py3-none-any.whl (27.4 kB view details)

Uploaded Python 3

File details

Details for the file kepler_echo-0.3.0.tar.gz.

File metadata

  • Download URL: kepler_echo-0.3.0.tar.gz
  • Upload date:
  • Size: 28.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.11.15

File hashes

Hashes for kepler_echo-0.3.0.tar.gz
Algorithm Hash digest
SHA256 841c229d2c9a65215f04b4f2276434910cbcae0371b1d1cc27e5da65c34da53a
MD5 915535ea64909f881de359ee38727476
BLAKE2b-256 3a26b7343b6e2945de76abb28e0c5486e224c419a0c0e7f9ab83cceecc6394ae

See more details on using hashes here.

File details

Details for the file kepler_echo-0.3.0-py3-none-any.whl.

File metadata

  • Download URL: kepler_echo-0.3.0-py3-none-any.whl
  • Upload date:
  • Size: 27.4 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.11.15

File hashes

Hashes for kepler_echo-0.3.0-py3-none-any.whl
Algorithm Hash digest
SHA256 46d79682263f948634fbdbbb199d74da333c68829afdb332a878a817fbacfbca
MD5 4e0934958c4f9dbb81d43714127d69cf
BLAKE2b-256 8797489a1e63cde6f687dadf9ebfaf06d765c9866d289f3a5de020184bd7b353

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page