Skip to main content

AlgoBench - 算法基准测试分析工具的 Python SDK 与 CLI

Project description

AlgoBench Python SDK

AlgoBench 是一个算法基准测试分析工具,提供完整的统计分析和决策评估流水线。

安装

pip install algobench

快速开始

Python 库

from algobench import analyze

result = analyze("data.csv", baseline="V1", compare="V2", metrics=["HPWL", "runtime"])

print(result.decision.label)        # "建议上线"
print(result.decision.status)       # "yes"
print(result.statistics["HPWL"].mean_imp)  # 5.07
print(result.statistics["HPWL"].p_value)   # 6.13e-13

命令行

# 完整分析
algobench analyze data.csv -b V1 -c V2 -m HPWL,runtime

# 单指标统计
algobench stats data.csv -b V1 -c V2 -m HPWL

# 数据质量检查
algobench quality data.csv

# 数据诊断
algobench diagnose data.csv

# CSV 格式验证
algobench validate data.csv

CSV 数据格式

Case,Baseline/HPWL,New/HPWL,Baseline/runtime,New/runtime
case1,1200,1100,50,45
case2,800,750,30,28
  • 用例列:CaseBenchmarkTestInstance
  • 指标列格式:算法名/指标名(使用 / 分隔符)
  • 元数据列:# 前缀(如 #Size
  • 参数列:p_ 前缀(如 p_mode

分析标准

模式 说明
exploratory 探索模式,识别更多潜在改进
standard 标准模式,平衡灵敏度和可靠性
strict 严格模式,用于发布或关键决策
result = analyze("data.csv", baseline="V1", compare="V2", criteria_id="strict")

决策状态

状态 含义
yes 建议上线
watch 观察期,需要更多数据
no 不建议上线
insufficient 数据不足

依赖

  • Python >= 3.10
  • numpy >= 1.24
  • scipy >= 1.10
  • pandas >= 2.0

许可证

AGPL-3.0-only

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

algobench_sdk-1.0.0.tar.gz (31.2 kB view details)

Uploaded Source

Built Distribution

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

algobench_sdk-1.0.0-py3-none-any.whl (41.9 kB view details)

Uploaded Python 3

File details

Details for the file algobench_sdk-1.0.0.tar.gz.

File metadata

  • Download URL: algobench_sdk-1.0.0.tar.gz
  • Upload date:
  • Size: 31.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.10.18

File hashes

Hashes for algobench_sdk-1.0.0.tar.gz
Algorithm Hash digest
SHA256 823d40cf93035cc162a67c41800755d4546eb4bd18b273d3605417b905832b47
MD5 316efabab737fda09c335bcfc3400401
BLAKE2b-256 62929d55398b80a1175406e401b5ef8a0617663c461e4d9bad8a6df116bbdb75

See more details on using hashes here.

File details

Details for the file algobench_sdk-1.0.0-py3-none-any.whl.

File metadata

  • Download URL: algobench_sdk-1.0.0-py3-none-any.whl
  • Upload date:
  • Size: 41.9 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.10.18

File hashes

Hashes for algobench_sdk-1.0.0-py3-none-any.whl
Algorithm Hash digest
SHA256 15e528607e74bc2d9dd72590882059b67a9f89c06bcffd3c3fa572e7ae6b9f27
MD5 b9fada5ebf37ab491bf73888c7100647
BLAKE2b-256 bc23e026394aa0029cbd85da27cbfa2626a19a1335d8d5f8d9367cf142e23057

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