Skip to main content

zlt-ai-tools

给 Python 用的智量通量化工具包。

一、能做什么

装上它,你的程序可以:

你想做的事 用哪个
算技术指标、跑自己写的公式 —— 通达信、同花顺等方言的公式直接编译执行,136 个内置函数 native.formula
做因子研究 —— 算因子、评有效性(信息系数、分层收益、换手率)、出可入库比对的结果 research
取行情 —— K 线、分时、财务、板块分类、排行榜 native.datafeed
查数据资产 —— 用自然语言找表、找字段、找证券 native.aiquery
联网检索 —— 搜新闻、研报、政策、论文 native.websearch
下单与查持仓(含仿真盘) trade
轻量取行情 —— 不落盘、只发 HTTP market_light

一句话:你写公式,它负责编译、取数、算出来,还能告诉你这个因子有没有用。

装完还会带上一批命令行程序(zlt-formulazlt-datafeedzlt-trade 等), 装在环境的 bin/Scripts 里,直接就能敲。

不需要装 Node.js、Go 或其他语言运行时。每个平台架构一个轮子,pip 自动挑。

二、先拿账号

除公式计算外的能力都要账号——行情、问数、检索都在服务端。

# 扫码登录,拿到 API Key 并存进本机凭据库
zlt-qrlogin -identity gateway-api-key -entry <给这把钥匙起个名>

跑起来会存一张二维码图片并告诉你路径,用微信扫它,确认之后钥匙自动落库。 钥匙不会打印在屏幕上,也不进日志和命令行参数——程序取用时从凭据库读, 人不需要看见它。

如果这个微信还没绑手机号,会让你补一次短信验证;照提示做即可。

只算公式、不取数的话,这一步可以跳过。

三、怎么用

算一条公式

from zlt_ai_tools import native

compiled = native.formula.compile(dialect="tdx",
                                  source="OUT:(CLOSE-MA(CLOSE,20))/MA(CLOSE,20);")
bars = native.datafeed.bars(symbol="SH600519", frequency="1d", limit=120)
result = native.formula.execute(handle=compiled["handle"], payload=bars["value"])
native.formula.free(handle=compiled["handle"])       # 用完还回去
print(result["value"][-1])                            # 最新一根的因子值

评一个因子有没有用

from zlt_ai_tools import research

report = research.run_factor_study(
    symbols=["SH600519", "SZ000001", "SH601318", "SZ000858",
             "SH600036", "SZ002415", "SH601899", "SZ300750"],
    start="2026-01-01", end="2026-06-30",
    dialect="tdx", source="OUT:(CLOSE-MA(CLOSE,20))/MA(CLOSE,20);",
    quantiles=5,
)
print(report.ic_mean)           # 因子与下一期收益的相关性
print(report.top_minus_bottom)  # 最高一层减最低一层的收益
print(report.turnover_mean)     # 换手率:太高就吃不到收益
for line in report.diagnostics:
    print(line)                 # 跳过的期、取不到的标的都在这

三个指标要一起看:相关性再高,分层不单调多半是被极端值带的; 前两项再好,换手率覆盖不了成本也是纸面收益。

看有哪些能力可以调

from zlt_ai_tools import native
print(native.version())    # 版本
print(native.actions())    # 全部动作,形如 "formula.compile"、"datafeed.bars"
print(native.transport)    # 这一次走的是共享库还是协进程

方法名按 域.动作 组织,native.<域>.<动作>() 直接调。actions() 是从原生件 自报的清单生成的——以它为准,文档可能滞后,那份清单不会。

四、出问题时

症状 去哪看
pip 说找不到匹配的发行版 手册「安装与平台覆盖」;确认 Python ≥ 3.8 与平台架构
调用报未授权 / 401 / 403 第二节重新扫码;钥匙可能过期或没绑到标识上
公式编译报错 手册的方言分卷:不同方言的函数与语义不完全一样
因子研究报「没有共同的期」 两侧的期标识不是一套写法;这不是「因子没用」,是数据没对齐
取数为空 手册「行情取数」:先确认代码写法与频率参数
命令行程序敲不动 确认装在虚拟环境里且已激活;轮子把命令装进环境的 bin/Scripts

手册在发布页的 manuals 包里(每个版本的 Release 附件),解压后打开 index.html。里面按能力线分卷:公式引擎、行情取数、在线搜索与问数、交易、 量化研究。

报问题时请附上 native.version()native.transport 的输出——这两样决定 了排障从哪开始:版本对不上和通道不同,症状可能一模一样而原因完全不同。


附:技术细节

这些不影响使用,排障或做集成时才需要。

  • 两条传输通道:优先走 C 共享库(进程内调用);musl 发行版(Alpine 等) 上共享库加载不了,自动改走协进程——同一份实现、同一个调用面,transport 告诉你这次走的哪条。
  • 包的构成:每个平台架构一个轮子,内含该平台的共享库或协进程程序, 外加同批命令行程序。行情轻客户端、交易客户端与量化研究能力包的源码已内联, 装它不需要另外解析依赖。
  • 聚合面自己不实现业务算法:因子研究的计算在 zlt_quantresearcher, 取数与公式由聚合面注入——能力线不反向依赖包装层。
  • 图表 SDK 不在本包公开面里:内含上游分支,公开分发要先过许可合规复核。

Release files for zlt-ai-tools 1.1.3

For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.

Built distributions (wheels)

Table of built distributions (wheels) for zlt-ai-tools 1.1.3
File
zlt_ai_tools-1.1.3-py3-none-win_arm64.whl Python 3 none Windows ARM64 Details
zlt_ai_tools-1.1.3-py3-none-win_amd64.whl Python 3 none Windows x86-64 Details
zlt_ai_tools-1.1.3-py3-none-musllinux_1_2_x86_64.whl Python 3 none Linux musl 1.2+ x86-64 Details
zlt_ai_tools-1.1.3-py3-none-musllinux_1_2_aarch64.whl Python 3 none Linux musl 1.2+ ARM64 Details
zlt_ai_tools-1.1.3-py3-none-manylinux_2_17_x86_64.whl Python 3 none Linux glibc 2.17+ x86-64 Details
zlt_ai_tools-1.1.3-py3-none-manylinux_2_17_aarch64.whl Python 3 none Linux glibc 2.17+ ARM64 Details

Total release size: 336.3 MB

Release files / zlt_ai_tools-1.1.3-py3-none-win_arm64.whl

Download URL zlt_ai_tools-1.1.3-py3-none-win_arm64.whl
Size 52.7 MB
Tags Python 3 Windows ARM64
SHA-256 checksum
How to use checksums
ddd2f5519853dc67a00c3b8dd5ba5db0089044b63c0a16acd77749377855312d
BLAKE2b-256 checksum
How to use checksums
6c8b7e8990678ded2304cb9807533506babfc600cc84d460cc84e7dad43cb113
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.14.3

Release files / zlt_ai_tools-1.1.3-py3-none-win_amd64.whl

Download URL zlt_ai_tools-1.1.3-py3-none-win_amd64.whl
Size 57.1 MB
Tags Python 3 Windows x86-64
SHA-256 checksum
How to use checksums
34aa77792a29818e6c61085a0ad08bad9f1940e28182ddf47252fc56d328b7ef
BLAKE2b-256 checksum
How to use checksums
707bb01de5e86647f78a223d75186afaa06af3d7bb7d171822904cea0a83b8ce
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.14.3

Release files / zlt_ai_tools-1.1.3-py3-none-musllinux_1_2_x86_64.whl

Download URL zlt_ai_tools-1.1.3-py3-none-musllinux_1_2_x86_64.whl
Size 60.9 MB
Tags Linux musl 1.2+ x86-64 Python 3
SHA-256 checksum
How to use checksums
5adbdc656feaa3465252c6a1cf8f171cdcd6fef46f46000c0a3f575c1f8a81bc
BLAKE2b-256 checksum
How to use checksums
58d493a688554d688dc52e295d5de8a14f9c7dac129c008b518ea133759352c7
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.14.3

Release files / zlt_ai_tools-1.1.3-py3-none-musllinux_1_2_aarch64.whl

Download URL zlt_ai_tools-1.1.3-py3-none-musllinux_1_2_aarch64.whl
Size 56.6 MB
Tags Linux musl 1.2+ ARM64 Python 3
SHA-256 checksum
How to use checksums
218df60f495e212cc180483ad09e5439f37ea6dd10dd7876bb10cfebd059220e
BLAKE2b-256 checksum
How to use checksums
13b9446b099b2a167c4badcda1c23e2060cace8f0a8eba60260c5c3a32ed0bb4
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.14.3

Release files / zlt_ai_tools-1.1.3-py3-none-manylinux_2_17_x86_64.whl

Download URL zlt_ai_tools-1.1.3-py3-none-manylinux_2_17_x86_64.whl
Size 56.4 MB
Tags Linux glibc 2.17+ x86-64 Python 3
SHA-256 checksum
How to use checksums
ed93c591d725338e246edf6fc85c5150144512aab75a68944a432baee722561c
BLAKE2b-256 checksum
How to use checksums
e59d4728da0264efeb67ee5c9c99f2d661dd0f62d8c2473bbf1d23b49fa85516
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.14.3

Release files / zlt_ai_tools-1.1.3-py3-none-manylinux_2_17_aarch64.whl

Download URL zlt_ai_tools-1.1.3-py3-none-manylinux_2_17_aarch64.whl
Size 52.5 MB
Tags Linux glibc 2.17+ ARM64 Python 3
SHA-256 checksum
How to use checksums
c9b087717aaa1744e0625eeffab6a55f2857877c258e0ba8096c176cad021a4e
BLAKE2b-256 checksum
How to use checksums
b968030a7197780b163be152bf92cd341b2ca2385dc89fc0c4100a96fcd801eb
Upload date
Uploaded using Trusted Publishing?
What is trusted publishing?
No
Uploaded via twine/7.0.0 CPython/3.14.3

Release history Release notifications | RSS feed

This release

1.1.3 This release

6 release files

1.1.2

6 release files

1.1.1

6 release files

0.1.0

6 release files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page