Skip to main content

wbt Python Package

Python API for the wbt Rust backtesting engine.

中文文档

Development Objectives

This Python subproject aims to provide a practical research-facing interface for weight-based backtesting while keeping the heavy computation in Rust.

Design priorities:

  1. Keep data input flexible for common research formats.
  2. Return analysis-friendly outputs as pandas objects.
  3. Preserve one consistent metric schema across stats outputs.
  4. Provide plotting utilities that work directly on backtest outputs.

Project Layout

This directory is an independent Python subproject.

python/
|-- pyproject.toml
|-- README.md
|-- scripts/
|-- tests/
`-- wbt/

The Rust crate remains one level up at ../Cargo.toml. maturin builds the extension module from there.

Installation And Local Setup

Requirements:

  • Rust toolchain
  • Python 3.10+
  • uv

Setup:

cd python
uv sync --extra dev
uv run maturin develop --release

Quick Start

import pandas as pd
from wbt import WeightBacktest

df = pd.DataFrame(
    {
        "dt": [
            "2024-01-02 09:01:00",
            "2024-01-02 09:02:00",
            "2024-01-02 09:03:00",
            "2024-01-02 09:04:00",
        ],
        "symbol": ["AAPL", "AAPL", "AAPL", "AAPL"],
        "weight": [0.5, 0.2, 0.0, -0.3],
        "price": [185.0, 186.0, 186.5, 184.5],
    }
)

wb = WeightBacktest(
    df,
    digits=2,
    fee_rate=0.0002,
    n_jobs=4,
    weight_type="ts",   # "ts" or "cs"
    yearly_days=252,
)

print("all:", wb.stats)
print("long:", wb.long_stats)
print("short:", wb.short_stats)

print(wb.daily_return.head())
print(wb.dailys.head())
print(wb.pairs.head())

print(wb.segment_stats("2024-01-01", "2024-12-31", kind="多空"))
print(wb.long_alpha_stats)

Accepted Inputs

The data argument accepts:

  • pandas.DataFrame
  • polars.DataFrame
  • polars.LazyFrame
  • file path as str or Path

Supported file formats from path input:

  • csv
  • parquet
  • feather
  • arrow

Required columns:

Column Type Meaning
dt datetime-like Bar end time
symbol str Instrument code
weight float Target position weight
price float Price used for return calculation

Notes:

  • Null values are not allowed.
  • weight is rounded by digits before backtest.

Main API Surface

Top-level imports (all reachable from import wbt):

from wbt import (
    # Backtest engine
    WeightBacktest, backtest,
    # Performance metrics (Rust-backed)
    daily_performance,
    top_drawdowns,
    rolling_daily_performance,
    cal_yearly_days,
    # Strategy utilities (pure Python)
    weights_simple_ensemble,
    cal_trade_price,
    log_strategy_info,
    # Reporting
    generate_backtest_report,
    # Test data
    mock_symbol_kline, mock_weights,
)

Primary class and helpers:

  • WeightBacktest(...): main backtest engine entry.
  • backtest(...): convenience wrapper returning a WeightBacktest.
  • daily_performance(returns, yearly_days=252): standalone metric utility on a daily-return array.
  • top_drawdowns(returns, top=10): top-N drawdown windows.
  • rolling_daily_performance(df, ret_col, window=252, min_periods=100, yearly_days=None): rolling-window daily performance.
  • cal_yearly_days(dts): infer yearly trading-day count from a date series.
  • weights_simple_ensemble(df, weight_cols, method="mean", only_long=False, **kwargs): ensemble multiple strategy weights (mean / vote / sum_clip).
  • cal_trade_price(df, digits=None, windows=(5, 10, 15, 20, 30, 60)): TWAP / VWAP and next-bar trade-price table grouped by symbol.
  • log_strategy_info(strategy, df): pretty-print per-symbol weight summaries via loguru.
  • generate_backtest_report(wb, output_path): render a self-contained HTML report.
  • mock_symbol_kline(...) / mock_weights(...): generators for quick experiments.

Core WeightBacktest properties and methods:

  • stats, long_stats, short_stats
  • daily_return, long_daily_return, short_daily_return
  • dailys, pairs
  • alpha, alpha_stats, bench_stats
  • segment_stats(sdt, edt, kind)
  • long_alpha_stats
  • get_symbol_daily(symbol), get_symbol_pairs(symbol)

Logging Note

cal_yearly_days and rolling_daily_performance emit warnings from Rust (e.g. short-span fallback) via the log crate. The package initializes pyo3-log at module load, so those warnings show up through Python's standard logging. If you use loguru, install an InterceptHandler once to route them into your loguru sinks.

Plotting Utilities

All plotting functions are single-purpose figures that consume a BacktestResult (from wb.to_result()) with zero data transformation — each field maps straight to a plotly trace. There are no composite (subplot) charts; the HTML report composes single figures into a CSS grid instead.

from wbt.plotting import (
    plot_colored_table,        # stats as a colored table
    plot_cumulative_returns,   # cumulative curves (voladj=True for vol-normalized)
    plot_daily_return_dist,    # daily-return histogram
    plot_drawdown,             # drawdown + cumulative (dual-axis single figure)
    plot_drawdowns_table,      # top-drawdowns detail table
    plot_key_trades,           # yearly best/worst key trades
    plot_monthly_heatmap,      # monthly-return heatmap
    plot_pairs_hold_dist,      # holding-bars distribution by direction
    plot_pairs_pnl_dist,       # pnl-ratio distribution by direction
    plot_rolling_metrics,      # rolling sharpe/return/vol over time (252d window)
    plot_segment_comparison,   # recent-1y vs full-sample metric table
    plot_stats_comparison,     # 多空/多头/空头/基准/超额 metric comparison table
    plot_symbol_returns,       # per-symbol cumulative returns
    plot_verdict,              # is_good_strategy verdict + yearly metrics
    plot_yearly_returns,       # yearly absolute vs excess returns (grouped bars)
)

Typical usage:

result = wb.to_result()

fig1 = plot_cumulative_returns(result, keys=["多空", "多头", "空头", "基准"])
fig2 = plot_cumulative_returns(result, voladj=True)   # vol-normalized
fig3 = plot_drawdown(result)
fig4 = plot_pairs_pnl_dist(result)

# Optional HTML export
html = plot_cumulative_returns(result, to_html=True)

# Full HTML report file (composes single figures into a tabbed CSS grid)
generate_backtest_report(df, "report.html")

Quality And Testing

Run checks from python/:

uv run pytest -v
uv run ruff format --check .
uv run ruff check . --no-fix
uv run basedpyright

Architecture Snapshot

repo-root/
|-- Cargo.toml
|-- src/
|   |-- lib.rs                       # pure Rust crate entry; Python bindings are off by default
|   |-- python.rs                    # PyO3 bindings (enabled by the python feature)
|   `-- core/
|       |-- cal_yearly_days.rs       # Rust core for cal_yearly_days
|       |-- daily_performance.rs
|       |-- rolling_daily_performance.rs
|       |-- top_drawdowns.rs
|       `-- ...                      # backtest engine internals
`-- python/
    `-- wbt/
        |-- __init__.py              # top-level exports
        |-- _df_convert.py           # pandas <-> Arrow IPC helpers
        |-- _wbt.pyi                 # Rust extension stubs
        |-- backtest.py              # WeightBacktest class
        |-- mock.py                  # mock_symbol_kline / mock_weights
        |-- top_drawdowns.py         # adapter for _wbt.top_drawdowns
        |-- utils/                   # adapters + pure-Python utilities
        |   |-- __init__.py
        |   |-- cal_yearly_days.py
        |   |-- rolling_daily_performance.py
        |   |-- weights_simple_ensemble.py
        |   |-- cal_trade_price.py
        |   `-- log_strategy_info.py
        |-- plotting/                # single-purpose plotly charts
        |   |-- __init__.py
        |   |-- _common.py
        |   |-- returns.py
        |   |-- risk.py
        |   |-- trades.py
        |   `-- overview.py
        `-- report/                  # HTML report + composite charts
            |-- __init__.py
            |-- _generator.py
            |-- _plot_backtest.py
            `-- html_builder.py

License

MIT

Download files

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

Source Distribution

wbt-0.5.0.tar.gz (2.2 MB view details)

Uploaded Source

Built Distributions

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

wbt-0.5.0-cp310-abi3-win_amd64.whl (21.4 MB view details)

Uploaded CPython 3.10+Windows x86-64

wbt-0.5.0-cp310-abi3-manylinux_2_28_x86_64.whl (19.2 MB view details)

Uploaded CPython 3.10+manylinux: glibc 2.28+ x86-64

wbt-0.5.0-cp310-abi3-manylinux_2_28_aarch64.whl (17.6 MB view details)

Uploaded CPython 3.10+manylinux: glibc 2.28+ ARM64

wbt-0.5.0-cp310-abi3-macosx_11_0_arm64.whl (17.3 MB view details)

Uploaded CPython 3.10+macOS 11.0+ ARM64

wbt-0.5.0-cp310-abi3-macosx_10_12_x86_64.whl (19.0 MB view details)

Uploaded CPython 3.10+macOS 10.12+ x86-64

File details

Details for the file wbt-0.5.0.tar.gz.

File metadata

  • Download URL: wbt-0.5.0.tar.gz
  • Upload date:
  • Size: 2.2 MB
  • Tags: Source
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for wbt-0.5.0.tar.gz
Algorithm Hash digest
SHA256 82d069de29a865f52c0395f6019ec13e5b3bfa4207bfde123091a68375cbba5d
MD5 5f123fc43d3f49e75ba3d59b6b6718e2
BLAKE2b-256 c01588b79a4035ff317c8ef8a71a06b3dd60f5d3dd7b911278212a96e7b1d4e1

See more details on using hashes here.

Provenance

The following attestation bundles were made for wbt-0.5.0.tar.gz:

Publisher: release.yml on zengbin93/wbt

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file wbt-0.5.0-cp310-abi3-win_amd64.whl.

File metadata

  • Download URL: wbt-0.5.0-cp310-abi3-win_amd64.whl
  • Upload date:
  • Size: 21.4 MB
  • Tags: CPython 3.10+, Windows x86-64
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for wbt-0.5.0-cp310-abi3-win_amd64.whl
Algorithm Hash digest
SHA256 26b6fc822ce571424500dd39750a4ae27da27a8dd786530478600d1adbfa9540
MD5 09474351e42318b630e58b3dc0d8d080
BLAKE2b-256 e28fb8098d09251dc3f5a4f2443a44d26f7012b9805d68a1853cdd5100f6f5fc

See more details on using hashes here.

Provenance

The following attestation bundles were made for wbt-0.5.0-cp310-abi3-win_amd64.whl:

Publisher: release.yml on zengbin93/wbt

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file wbt-0.5.0-cp310-abi3-manylinux_2_28_x86_64.whl.

File metadata

  • Download URL: wbt-0.5.0-cp310-abi3-manylinux_2_28_x86_64.whl
  • Upload date:
  • Size: 19.2 MB
  • Tags: CPython 3.10+, manylinux: glibc 2.28+ x86-64
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for wbt-0.5.0-cp310-abi3-manylinux_2_28_x86_64.whl
Algorithm Hash digest
SHA256 eb476f1042c1a237b7d7e50e96cb808ceb1b8417c979768012c76c89c088be3e
MD5 983e79822babef1068dcacb277a639d7
BLAKE2b-256 433a1746a75860bc0161cad166fd4a5398be5a19c884ce92b3ae440d092895f5

See more details on using hashes here.

Provenance

The following attestation bundles were made for wbt-0.5.0-cp310-abi3-manylinux_2_28_x86_64.whl:

Publisher: release.yml on zengbin93/wbt

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file wbt-0.5.0-cp310-abi3-manylinux_2_28_aarch64.whl.

File metadata

File hashes

Hashes for wbt-0.5.0-cp310-abi3-manylinux_2_28_aarch64.whl
Algorithm Hash digest
SHA256 482144de5105b53b70fc31be5747686c7e3747852c58e390139636b537b4c81f
MD5 661b1bbb5aa94f6588ea62c1cbfb3e59
BLAKE2b-256 95565e6e0bb415c15afdc3c9c6670a17ea03417e037dd540d405927dfddfe546

See more details on using hashes here.

Provenance

The following attestation bundles were made for wbt-0.5.0-cp310-abi3-manylinux_2_28_aarch64.whl:

Publisher: release.yml on zengbin93/wbt

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file wbt-0.5.0-cp310-abi3-macosx_11_0_arm64.whl.

File metadata

  • Download URL: wbt-0.5.0-cp310-abi3-macosx_11_0_arm64.whl
  • Upload date:
  • Size: 17.3 MB
  • Tags: CPython 3.10+, macOS 11.0+ ARM64
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for wbt-0.5.0-cp310-abi3-macosx_11_0_arm64.whl
Algorithm Hash digest
SHA256 8b25f35c7492046a9f4cc0f89ddcb193c9fd387df7a7481cd38aa1fa906b0ae8
MD5 40ae84c5320d30107566508baea545e5
BLAKE2b-256 ce3424823ea2854db14ac725a406d9a2d707b0768baeaed1d1a37bb8636830fa

See more details on using hashes here.

Provenance

The following attestation bundles were made for wbt-0.5.0-cp310-abi3-macosx_11_0_arm64.whl:

Publisher: release.yml on zengbin93/wbt

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

File details

Details for the file wbt-0.5.0-cp310-abi3-macosx_10_12_x86_64.whl.

File metadata

  • Download URL: wbt-0.5.0-cp310-abi3-macosx_10_12_x86_64.whl
  • Upload date:
  • Size: 19.0 MB
  • Tags: CPython 3.10+, macOS 10.12+ x86-64
  • Uploaded using Trusted Publishing? Yes
  • Uploaded via: twine/6.1.0 CPython/3.13.12

File hashes

Hashes for wbt-0.5.0-cp310-abi3-macosx_10_12_x86_64.whl
Algorithm Hash digest
SHA256 d73554620cc8faa3e1bd84d08d230b7b116ae2dc7bfd9f6a24e332a4715d2955
MD5 39fa9de09b9cc05ed35620b6cc6df9b6
BLAKE2b-256 cebbffaee0e89c32c2bc858cf3bf7e8548106b10cf8d23b622f1fc507c46bb2b

See more details on using hashes here.

Provenance

The following attestation bundles were made for wbt-0.5.0-cp310-abi3-macosx_10_12_x86_64.whl:

Publisher: release.yml on zengbin93/wbt

Attestations: Values shown here reflect the state when the release was signed and may no longer be current.

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