python-quanta 0.8.0

A quantitative analysis project for financial data.

quanta

Summary

A local-first quantitative research framework designed for:
    -- providing specialized financial analysis modules and functions.
    -- abstracting database connectivity to streamline data access for researchers.
    -- deep integration with JoinQuant (JQData) for professional-grade data support.
    -- supporting industry-standard factor models (e.g., Barra - Completed Style Factors).

How Install

pip install python-quanta -i https://pypi.org/simple

How Initial Local Settings

I new docoument like:
    your project
    |-- .env                                 <-- build by user
         |-- account.yaml                    <-- build by user if you have
         |-- libs.yaml                       <-- must build

II if having join quant account, create file <account.yaml>:
    account.yaml
    ----
    joinquant:
        username: 'your_user_name'
        password: 'your_password'

III set up your database env, create file <libs.yaml>:
    libs.yaml
    ----
    db:
        recommand_settings: DuckDB            # recommand database type
        DuckDB:
            recommand_settings:
                path: 'E:/ProgramData/DuckDB' # path of database file
                database: Locals              # name of database file
                schema: jq_data               # schema name

For technical details on the underlying database abstraction and supported engines, see:
[Database Abstraction Layer](src/quanta/libs/db/README.md)

For a detailed explanation of all configuration fields, see:
[Configuration Definitions](src/quanta/config/README.md)

Why python-quanta

I provide local database construct automatically
    all you need is just running:
        -- quanta.data.daily()
    including:
        -- trading data: stock, index(not in CSI), etf
        -- finance data: stock

    For details on data provider integration and table logic, see:
    [Data Management](src/quanta/data/README.md)

II specialized financial analytical extensions (based on pandas.DataFrame)
    -- DataFrame.gen
                    .group()                 # quick-grouping based on multi-factor theory
                    .portfolio()             # get returns of each group generated by function above
    -- DataFrame.rollings
                    .sth()                   # specialized rolling, e.g., "get 3 max volume in 21 days"
    -- DataFrame.stats
                    .neutral()               # OLS or factor neutralization (e.g., .stats.neutral)

    For a full reference of available accessors and metrics, see:
    [Analytical Extensions](src/quanta/libs/_pandas/README.md)

III more powerful functions with database
    if user have join quant account, python-quanta will be more powerful:
    -- DataFrame.f.listing                   # filter portfolio by listed date
                   not_st                    # filter portfolio by ST status
                   tradestatus               # filter portfolio by trade status
                   index_members             # get index constituents (e.g., '300', '500')
                   label                     # get industry classifications (e.g., "swl1_code")
                   ic                        # Information Coefficient (IC)
                   ir                        # Information Ratio (IR, annual)
                   port / test               # rapid group return calculation and backtesting

    For a deep dive into research interfaces and Pandas extensions, see: 
    [Research Flow Layer](src/quanta/libs/_flow/README.md) --> _extra_pandas
    
IV high-efficiency research workflow (quanta.flow)
    The `flow` module provides a high-level abstraction for factor/strategy research:
    -- quanta.flow.astock("key")             # unified interface for price and financial data
    -- quanta.flow.astock._help              # metadata lookup for tables, columns, and comments
    -- quanta.flow.astock.help("keyword")    # fuzzy search within metadata (e.g., "oper" for operating data)

    For a deep dive into research interfaces and Pandas extensions, see:
    [Research Flow Layer](src/quanta/libs/_flow/README.md) --> _extra_pandas

V standard factor library (faclib)
    Benchmark factors for risk modeling and strategy comparison:
    -- Barra USA4 Style Factors (Completed): 
       Size, Beta, Momentum, Residual Volatility, Liquidity, Earnings Yield, Growth, Leverage, etc.
    -- Alpha 101 (Roadmap)

Notices

welcome for advise! ^_^
if you want data for testing python-quanta, please mail: [e-mail]

---

quanta (中文版)

Summary

本地优先的量化研究框架, 旨在:
    -- 提供针对金融场景优化的专业分析模块与函数.
    -- 抽象化数据库连接细节, 使研究员能够高效获取所需数据.
    -- 深度集成聚宽 (JQData) API, 提供强大的专业数据支持.
    -- 支持行业标准因子模型 (如 Barra 风险模型 - 已完成风格因子).

How Install

pip install python-quanta -i https://pypi.org/simple

How Initial Local Settings

I 新建配置文件:
    你的项目目录/
    |-- .env                                 <-- 环境变量配置文件
         |-- account.yaml                    <-- 聚宽账户凭据 (可选)
         |-- libs.yaml                       <-- 数据库配置 (必填)

II 若有聚宽账户, 创建 <account.yaml>:
    account.yaml
    ----
    joinquant:
        username: 'your_user_name'
        password: 'your_password'

III 配置数据库环境 <libs.yaml>:
    libs.yaml
    ----
    db:
        recommand_settings: DuckDB            # 推荐使用的存储引擎
        DuckDB:
            recommand_settings:
                path: 'E:/ProgramData/DuckDB' # 本地数据库存放路径
                database: Locals              # 数据库文件名
                schema: jq_data               # Schema 名称

有关底层数据库抽象及支持引擎的技术详情, 请参阅:
[数据库抽象层详解](src/quanta/libs/db/README.md)

有关所有配置字段的详细定义说明, 请参阅:
[配置定义详解](src/quanta/config/README.md)

Why python-quanta

I 提供自动化本地数据库构建
    仅需运行以下指令即可全自动同步 ETL:
        -- quanta.data.daily()
    包含数据:
        -- 交易数据: 股票, 指数(非中证系列), ETF.
        -- 财务数据: 全量股票财务报表.

    有关数据供应商集成及表逻辑的详细说明, 请参阅:
    [数据管理详解](src/quanta/data/README.md)

II 专业金融分析扩展 (基于 pandas.DataFrame)
    -- DataFrame.gen
                    .group()                 # 基于多因子理论的高效分组函数
                    .portfolio()             # 计算基于分组结果的投资组合收益
    -- DataFrame.rollings
                    .sth()                   # 专用滚动窗口函数(如: "获取21日内成交量最大的3个交易日")
    -- DataFrame.stats
                    .neutral()               # 因子中性化处理及截面 OLS 回归

    有关可用访问器及指标的完整参考, 请参阅:
    [专业分析扩展详解](src/quanta/libs/_pandas/README.md)

III 数据库增强型高级功能
    需配合聚宽账户以实现更强大的分析能力:
    -- DataFrame.f.listing                   # 根据上市日期过滤
                   not_st                    # 剔除 ST/退市风险警示股票
                   tradestatus               # 根据交易状态过滤
                   index_members             # 获取指数成分股(如 '300', '500')
                   label                     # 获取行业分类信息(如申万一级)
                   ic                        # 因子 IC (Information Coefficient) 计算
                   ir                        # 因子 IR (Information Ratio, 年化) 计算
                   port / test               # 快速计算分组收益及策略回测

    有关研究接口及 Pandas 扩展功能的深入说明, 请参阅:
    [研究流层详解](src/quanta/libs/_flow/README.md)  --> _extra_pandas

IV 高效研究流工具 (quanta.flow)
    `flow` 模块为因子与策略研发提供高层抽象:
    -- quanta.flow.astock("key")             # 获取价格或财务数据的统一接口
    -- quanta.flow.astock._help              # 快速查询数据库表, 字段及注释信息
    -- quanta.flow.astock.help("key")        # 模糊搜索元数据(如搜索 "oper" 获取营业相关字段)

    有关研究接口及 Pandas 扩展功能的深入说明, 请参阅:
    [研究流层详解](src/quanta/libs/_flow/README.md)  --> _main

V 标准因子库 (faclib)
    持续集成行业标准因子以供对比与测试:
    -- Barra USA4 风格因子 (已完成): 
       包含市值、贝塔、动量、残差波动率、流动性、盈利、成长、杠杆等 10 大风格因子.
    -- Alpha 101 系列因子 (发展路线).

Notices

欢迎任何建议与反馈! ^_^
如需测试数据或技术咨询, 请联系: [e-mail]