[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"similar-numenta--nupic-legacy":3,"tool-numenta--nupic-legacy":64},[4,17,27,35,43,56],{"id":5,"name":6,"github_repo":7,"description_zh":8,"stars":9,"difficulty_score":10,"last_commit_at":11,"category_tags":12,"status":16},3808,"stable-diffusion-webui","AUTOMATIC1111\u002Fstable-diffusion-webui","stable-diffusion-webui 是一个基于 Gradio 构建的网页版操作界面，旨在让用户能够轻松地在本地运行和使用强大的 Stable Diffusion 图像生成模型。它解决了原始模型依赖命令行、操作门槛高且功能分散的痛点，将复杂的 AI 绘图流程整合进一个直观易用的图形化平台。\n\n无论是希望快速上手的普通创作者、需要精细控制画面细节的设计师，还是想要深入探索模型潜力的开发者与研究人员，都能从中获益。其核心亮点在于极高的功能丰富度：不仅支持文生图、图生图、局部重绘（Inpainting）和外绘（Outpainting）等基础模式，还独创了注意力机制调整、提示词矩阵、负向提示词以及“高清修复”等高级功能。此外，它内置了 GFPGAN 和 CodeFormer 等人脸修复工具，支持多种神经网络放大算法，并允许用户通过插件系统无限扩展能力。即使是显存有限的设备，stable-diffusion-webui 也提供了相应的优化选项，让高质量的 AI 艺术创作变得触手可及。",162132,3,"2026-04-05T11:01:52",[13,14,15],"开发框架","图像","Agent","ready",{"id":18,"name":19,"github_repo":20,"description_zh":21,"stars":22,"difficulty_score":23,"last_commit_at":24,"category_tags":25,"status":16},1381,"everything-claude-code","affaan-m\u002Feverything-claude-code","everything-claude-code 是一套专为 AI 编程助手（如 Claude Code、Codex、Cursor 等）打造的高性能优化系统。它不仅仅是一组配置文件，而是一个经过长期实战打磨的完整框架，旨在解决 AI 代理在实际开发中面临的效率低下、记忆丢失、安全隐患及缺乏持续学习能力等核心痛点。\n\n通过引入技能模块化、直觉增强、记忆持久化机制以及内置的安全扫描功能，everything-claude-code 能显著提升 AI 在复杂任务中的表现，帮助开发者构建更稳定、更智能的生产级 AI 代理。其独特的“研究优先”开发理念和针对 Token 消耗的优化策略，使得模型响应更快、成本更低，同时有效防御潜在的攻击向量。\n\n这套工具特别适合软件开发者、AI 研究人员以及希望深度定制 AI 工作流的技术团队使用。无论您是在构建大型代码库，还是需要 AI 协助进行安全审计与自动化测试，everything-claude-code 都能提供强大的底层支持。作为一个曾荣获 Anthropic 黑客大奖的开源项目，它融合了多语言支持与丰富的实战钩子（hooks），让 AI 真正成长为懂上",138956,2,"2026-04-05T11:33:21",[13,15,26],"语言模型",{"id":28,"name":29,"github_repo":30,"description_zh":31,"stars":32,"difficulty_score":23,"last_commit_at":33,"category_tags":34,"status":16},2271,"ComfyUI","Comfy-Org\u002FComfyUI","ComfyUI 是一款功能强大且高度模块化的视觉 AI 引擎，专为设计和执行复杂的 Stable Diffusion 图像生成流程而打造。它摒弃了传统的代码编写模式，采用直观的节点式流程图界面，让用户通过连接不同的功能模块即可构建个性化的生成管线。\n\n这一设计巧妙解决了高级 AI 绘图工作流配置复杂、灵活性不足的痛点。用户无需具备编程背景，也能自由组合模型、调整参数并实时预览效果，轻松实现从基础文生图到多步骤高清修复等各类复杂任务。ComfyUI 拥有极佳的兼容性，不仅支持 Windows、macOS 和 Linux 全平台，还广泛适配 NVIDIA、AMD、Intel 及苹果 Silicon 等多种硬件架构，并率先支持 SDXL、Flux、SD3 等前沿模型。\n\n无论是希望深入探索算法潜力的研究人员和开发者，还是追求极致创作自由度的设计师与资深 AI 绘画爱好者，ComfyUI 都能提供强大的支持。其独特的模块化架构允许社区不断扩展新功能，使其成为当前最灵活、生态最丰富的开源扩散模型工具之一，帮助用户将创意高效转化为现实。",107662,"2026-04-03T11:11:01",[13,14,15],{"id":36,"name":37,"github_repo":38,"description_zh":39,"stars":40,"difficulty_score":23,"last_commit_at":41,"category_tags":42,"status":16},3704,"NextChat","ChatGPTNextWeb\u002FNextChat","NextChat 是一款轻量且极速的 AI 助手，旨在为用户提供流畅、跨平台的大模型交互体验。它完美解决了用户在多设备间切换时难以保持对话连续性，以及面对众多 AI 模型不知如何统一管理的痛点。无论是日常办公、学习辅助还是创意激发，NextChat 都能让用户随时随地通过网页、iOS、Android、Windows、MacOS 或 Linux 端无缝接入智能服务。\n\n这款工具非常适合普通用户、学生、职场人士以及需要私有化部署的企业团队使用。对于开发者而言，它也提供了便捷的自托管方案，支持一键部署到 Vercel 或 Zeabur 等平台。\n\nNextChat 的核心亮点在于其广泛的模型兼容性，原生支持 Claude、DeepSeek、GPT-4 及 Gemini Pro 等主流大模型，让用户在一个界面即可自由切换不同 AI 能力。此外，它还率先支持 MCP（Model Context Protocol）协议，增强了上下文处理能力。针对企业用户，NextChat 提供专业版解决方案，具备品牌定制、细粒度权限控制、内部知识库整合及安全审计等功能，满足公司对数据隐私和个性化管理的高标准要求。",87618,"2026-04-05T07:20:52",[13,26],{"id":44,"name":45,"github_repo":46,"description_zh":47,"stars":48,"difficulty_score":23,"last_commit_at":49,"category_tags":50,"status":16},2268,"ML-For-Beginners","microsoft\u002FML-For-Beginners","ML-For-Beginners 是由微软推出的一套系统化机器学习入门课程，旨在帮助零基础用户轻松掌握经典机器学习知识。这套课程将学习路径规划为 12 周，包含 26 节精炼课程和 52 道配套测验，内容涵盖从基础概念到实际应用的完整流程，有效解决了初学者面对庞大知识体系时无从下手、缺乏结构化指导的痛点。\n\n无论是希望转型的开发者、需要补充算法背景的研究人员，还是对人工智能充满好奇的普通爱好者，都能从中受益。课程不仅提供了清晰的理论讲解，还强调动手实践，让用户在循序渐进中建立扎实的技能基础。其独特的亮点在于强大的多语言支持，通过自动化机制提供了包括简体中文在内的 50 多种语言版本，极大地降低了全球不同背景用户的学习门槛。此外，项目采用开源协作模式，社区活跃且内容持续更新，确保学习者能获取前沿且准确的技术资讯。如果你正寻找一条清晰、友好且专业的机器学习入门之路，ML-For-Beginners 将是理想的起点。",84991,"2026-04-05T10:45:23",[14,51,52,53,15,54,26,13,55],"数据工具","视频","插件","其他","音频",{"id":57,"name":58,"github_repo":59,"description_zh":60,"stars":61,"difficulty_score":10,"last_commit_at":62,"category_tags":63,"status":16},3128,"ragflow","infiniflow\u002Fragflow","RAGFlow 是一款领先的开源检索增强生成（RAG）引擎，旨在为大语言模型构建更精准、可靠的上下文层。它巧妙地将前沿的 RAG 技术与智能体（Agent）能力相结合，不仅支持从各类文档中高效提取知识，还能让模型基于这些知识进行逻辑推理和任务执行。\n\n在大模型应用中，幻觉问题和知识滞后是常见痛点。RAGFlow 通过深度解析复杂文档结构（如表格、图表及混合排版），显著提升了信息检索的准确度，从而有效减少模型“胡编乱造”的现象，确保回答既有据可依又具备时效性。其内置的智能体机制更进一步，使系统不仅能回答问题，还能自主规划步骤解决复杂问题。\n\n这款工具特别适合开发者、企业技术团队以及 AI 研究人员使用。无论是希望快速搭建私有知识库问答系统，还是致力于探索大模型在垂直领域落地的创新者，都能从中受益。RAGFlow 提供了可视化的工作流编排界面和灵活的 API 接口，既降低了非算法背景用户的上手门槛，也满足了专业开发者对系统深度定制的需求。作为基于 Apache 2.0 协议开源的项目，它正成为连接通用大模型与行业专有知识之间的重要桥梁。",77062,"2026-04-04T04:44:48",[15,14,13,26,54],{"id":65,"github_repo":66,"name":67,"description_en":68,"description_zh":69,"ai_summary_zh":69,"readme_en":70,"readme_zh":71,"quickstart_zh":72,"use_case_zh":73,"hero_image_url":74,"owner_login":75,"owner_name":76,"owner_avatar_url":77,"owner_bio":78,"owner_company":79,"owner_location":79,"owner_email":79,"owner_twitter":79,"owner_website":80,"owner_url":81,"languages":82,"stars":114,"forks":115,"last_commit_at":116,"license":117,"difficulty_score":118,"env_os":119,"env_gpu":120,"env_ram":120,"env_deps":121,"category_tags":130,"github_topics":131,"view_count":118,"oss_zip_url":79,"oss_zip_packed_at":79,"status":16,"created_at":136,"updated_at":137,"faqs":138,"releases":166},748,"numenta\u002Fnupic-legacy","nupic-legacy","Numenta Platform for Intelligent Computing is an implementation of Hierarchical Temporal Memory (HTM), a theory of intelligence based strictly on the neuroscience of the neocortex.","nupic-legacy 是一个基于大脑新皮层神经科学理论的机器学习平台，核心实现了层级时序记忆（HTM）算法。它利用时间连续的机制来存储和回忆时空模式，特别适合解决流式数据的异常检测与预测问题。\n\n这个平台为人工智能研究者和开发者提供了一个探索类脑计算的窗口。不同于传统深度学习，nupic-legacy 强调在少量数据下持续学习的能力，非常适合需要处理时序动态变化的场景。\n\n值得注意的是，该项目目前已进入维护模式，官方仅保留关键 Bug 修复和研究支持功能。如果你正在复现旧版 HTM 项目，或者对神经形态计算感兴趣，nupic-legacy 仍是重要的参考资源。使用前请确保开发环境支持 Python 2.7 及 C++ 编译工具链。","# \u003Cimg src=\"http:\u002F\u002Fnumenta.org\u002F87b23beb8a4b7dea7d88099bfb28d182.svg\" alt=\"NuPIC Logo\" width=100\u002F> NuPIC\n\nAs of September 2023 this repository contains code from legacy Hierarchical Temporal Memory (HTM) Numenta projects that have been in maintenance mode for several years.\n\n## Numenta Platform for Intelligent Computing\n\nThe Numenta Platform for Intelligent Computing (**NuPIC**) is a machine intelligence platform that implements the [HTM learning algorithms](https:\u002F\u002Fnumenta.com\u002Fresources\u002Fpapers-videos-and-more\u002F). HTM is a detailed computational theory of the neocortex. At the core of HTM are time-based continuous learning algorithms that store and recall spatial and temporal patterns. NuPIC is suited to a variety of problems, particularly anomaly detection and prediction of streaming data sources. For more information, see [numenta.org](http:\u002F\u002Fnumenta.org) or the [NuPIC Forum](https:\u002F\u002Fdiscourse.numenta.org\u002Fc\u002Fnupic).\n\nFor usage guides, quick starts, and API documentation, see \u003Chttp:\u002F\u002Fnupic.docs.numenta.org\u002F>.\n\n## This project is in Maintenance Mode\n\nWe plan to do minor releases only, and limit changes in NuPIC and NuPIC Core to:\n\n- Fixing critical bugs.\n- Features needed to support ongoing research.\n\n## Installing NuPIC\n\nNuPIC binaries are available for:\n\n- Linux x86 64bit\n- OS X 10.9\n- OS X 10.10\n- Windows 64bit\n\n### Dependencies\n\nThe following dependencies are required to install NuPIC on all operating systems.\n\n- [Python 2.7](https:\u002F\u002Fwww.python.org\u002F)\n- [pip](https:\u002F\u002Fpip.pypa.io\u002Fen\u002Fstable\u002Finstalling\u002F)>=8.1.2\n- [setuptools](https:\u002F\u002Fsetuptools.readthedocs.io)>=25.2.0\n- [wheel](http:\u002F\u002Fpythonwheels.com)>=0.29.0\n- [numpy](http:\u002F\u002Fwww.numpy.org\u002F)\n- C++ 11 compiler like [gcc](https:\u002F\u002Fgcc.gnu.org\u002F) (4.8+) or [clang](http:\u002F\u002Fclang.llvm.org\u002F)\n\nAdditional OS X requirements:\n\n- [Xcode command line tools](https:\u002F\u002Fdeveloper.apple.com\u002Flibrary\u002Fios\u002Ftechnotes\u002Ftn2339\u002F_index.html)\n\n### Install\n\nRun the following to install NuPIC:\n\n    pip install nupic\n\n### Test\n\n    # From the root of the repo:\n    py.test tests\u002Funit\n\n### _Having problems?_\n\n- You may need to use the `--user` flag for the commands above to install in a non-system location (depends on your environment). Alternatively, you can execute the `pip` commands with `sudo` (not recommended).\n- You may need to add the `--use-wheel` option if you have an older pip version (wheels are now the default binary package format for pip).\n\nFor any other installation issues, please see our [search our forums](https:\u002F\u002Fdiscourse.numenta.org\u002Fsearch?q=tag%3Ainstallation%20category%3A10) (post questions there). You can report bugs at https:\u002F\u002Fgithub.com\u002Fnumenta\u002Fnupic\u002Fissues.\n\nLive Community Chat: [![Gitter](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fgitter-join_chat-blue.svg?style=flat)](https:\u002F\u002Fgitter.im\u002Fnumenta\u002Fpublic?utm_source=badge)\n\n### Installing NuPIC From Source\n\nTo install from local source code, run from the repository root:\n\n    pip install .\n\nUse the optional `-e` argument for a developer install.\n\nIf you want to build the dependent `nupic.bindings` from source, you should build and install from [`nupic.core`](https:\u002F\u002Fgithub.com\u002Fnumenta\u002Fnupic.core) prior to installing nupic (since a PyPI release will be installed if `nupic.bindings` isn't yet installed).\n\n- Build:\n[![Build Status](https:\u002F\u002Ftravis-ci.org\u002Fnumenta\u002Fnupic.png?branch=master)](https:\u002F\u002Ftravis-ci.org\u002Fnumenta\u002Fnupic)\n[![AppVeyor Status](https:\u002F\u002Fci.appveyor.com\u002Fapi\u002Fprojects\u002Fstatus\u002F4toemh0qtr21mk6b\u002Fbranch\u002Fmaster?svg=true)](https:\u002F\u002Fci.appveyor.com\u002Fproject\u002Fnumenta-ci\u002Fnupic\u002Fbranch\u002Fmaster)\n[![CircleCI](https:\u002F\u002Fcircleci.com\u002Fgh\u002Fnumenta\u002Fnupic.svg?style=svg)](https:\u002F\u002Fcircleci.com\u002Fgh\u002Fnumenta\u002Fnupic)\n- To cite this codebase: [![DOI](https:\u002F\u002Fzenodo.org\u002Fbadge\u002F19461\u002Fnumenta\u002Fnupic.svg)](https:\u002F\u002Fzenodo.org\u002Fbadge\u002Flatestdoi\u002F19461\u002Fnumenta\u002Fnupic)\n","# \u003Cimg src=\"http:\u002F\u002Fnumenta.org\u002F87b23beb8a4b7dea7d88099bfb28d182.svg\" alt=\"NuPIC Logo\" width=100\u002F> NuPIC\n\n截至 2023 年 9 月，此仓库包含来自长期处于维护模式的遗留层级时序记忆 (Hierarchical Temporal Memory, HTM) Numenta 项目的代码。\n\n## Numenta 智能计算平台\n\nNumenta 智能计算平台（**NuPIC**）是一个实现 [HTM learning algorithms](https:\u002F\u002Fnumenta.com\u002Fresources\u002Fpapers-videos-and-more\u002F) 的机器智能平台。HTM 是新皮层的一种详细计算理论。HTM 的核心是基于时间的连续学习算法，用于存储和回忆空间及时间模式。NuPIC 适用于各种问题，特别是异常检测和流数据源的预测。更多信息请见 [numenta.org](http:\u002F\u002Fnumenta.org) 或 [NuPIC Forum](https:\u002F\u002Fdiscourse.numenta.org\u002Fc\u002Fnupic)。\n\n有关使用指南、快速入门和 API 文档，请访问 \u003Chttp:\u002F\u002Fnupic.docs.numenta.org\u002F>。\n\n## 本项目处于维护模式\n\n我们计划仅进行小版本发布，并限制 NuPIC 和 NuPIC Core 的变更包括：\n\n- 修复关键错误。\n- 支持持续研究所需的功能。\n\n## 安装 NuPIC\n\nNuPIC 二进制文件适用于以下系统：\n\n- Linux x86 64 位\n- OS X 10.9\n- OS X 10.10\n- Windows 64 位\n\n### 依赖项\n\n在所有操作系统上安装 NuPIC 需要以下依赖项。\n\n- [Python 2.7](https:\u002F\u002Fwww.python.org\u002F)\n- [pip](https:\u002F\u002Fpip.pypa.io\u002Fen\u002Fstable\u002Finstalling\u002F)>=8.1.2\n- [setuptools](https:\u002F\u002Fsetuptools.readthedocs.io)>=25.2.0\n- [wheel](http:\u002F\u002Fpythonwheels.com)>=0.29.0\n- [numpy](http:\u002F\u002Fwww.numpy.org\u002F)\n- C++ 11 编译器如 [gcc](https:\u002F\u002Fgcc.gnu.org\u002F) (4.8+) 或 [clang](http:\u002F\u002Fclang.llvm.org\u002F)\n\n额外的 OS X 要求：\n\n- [Xcode command line tools](https:\u002F\u002Fdeveloper.apple.com\u002Flibrary\u002Fios\u002Ftechnotes\u002Ftn2339\u002F_index.html)\n\n### 安装\n\n运行以下内容以安装 NuPIC：\n\n    pip install nupic\n\n### 测试\n\n    # From the root of the repo:\n    py.test tests\u002Funit\n\n### _遇到问题？_\n\n- 您可能需要使用上述命令的 `--user` 标志在非系统位置安装（取决于您的环境）。或者，您可以使用 `sudo` 执行 `pip` 命令（不推荐）。\n- 如果您使用的是旧版 pip，可能需要添加 `--use-wheel` 选项（wheels 现在默认是 pip 的二进制包格式）。\n\n对于任何其他安装问题，请参阅我们的 [search our forums](https:\u002F\u002Fdiscourse.numenta.org\u002Fsearch?q=tag%3Ainstallation%20category%3A10)（在那里发帖提问）。您可以在 https:\u002F\u002Fgithub.com\u002Fnumenta\u002Fnupic\u002Fissues 报告错误。\n\n实时社区聊天：[![Gitter](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fgitter-join_chat-blue.svg?style=flat)](https:\u002F\u002Fgitter.im\u002Fnumenta\u002Fpublic?utm_source=badge)\n\n### 从源代码安装 NuPIC\n\n要从本地源代码安装，请在仓库根目录下运行：\n\n    pip install .\n\n使用可选的 `-e` 参数进行开发者安装。\n\n如果您想从源代码构建依赖项 `nupic.bindings`，您应该先构建并安装 [`nupic.core`](https:\u002F\u002Fgithub.com\u002Fnumenta\u002Fnupic.core)，然后再安装 nupic（因为如果 `nupic.bindings` 尚未安装，将安装 PyPI 版本）。\n\n- 构建：\n[![Build Status](https:\u002F\u002Ftravis-ci.org\u002Fnumenta\u002Fnupic.png?branch=master)](https:\u002F\u002Ftravis-ci.org\u002Fnumenta\u002Fnupic)\n[![AppVeyor Status](https:\u002F\u002Fci.appveyor.com\u002Fapi\u002Fprojects\u002Fstatus\u002F4toemh0qtr21mk6b\u002Fbranch\u002Fmaster?svg=true)](https:\u002F\u002Fci.appveyor.com\u002Fproject\u002Fnumenta-ci\u002Fnupic\u002Fbranch\u002Fmaster)\n[![CircleCI](https:\u002F\u002Fcircleci.com\u002Fgh\u002Fnumenta\u002Fnupic.svg?style=svg)](https:\u002F\u002Fcircleci.com\u002Fgh\u002Fnumenta\u002Fnupic)\n- 引用此代码库：[![DOI](https:\u002F\u002Fzenodo.org\u002Fbadge\u002F19461\u002Fnumenta\u002Fnupic.svg)](https:\u002F\u002Fzenodo.org\u002Fbadge\u002Flatestdoi\u002F19461\u002Fnumenta\u002Fnupic)","# NuPIC (Legacy) 快速上手指南\n\nNuPIC (Numenta Platform for Intelligent Computing) 是一个基于 HTM（层次时序记忆）理论的机器智能平台，主要用于异常检测和流数据预测。**注意：自 2023 年 9 月起，该项目已处于维护模式，仅进行关键 Bug 修复和支持研究的功能更新。**\n\n## 环境准备\n\n### 系统要求\n支持以下操作系统：\n- Linux x86_64\n- macOS 10.9 \u002F 10.10\n- Windows 64bit\n\n### 前置依赖\n所有操作系统均需满足以下条件：\n- **Python 2.7** (重要：需确保使用 Python 2 环境)\n- [pip](https:\u002F\u002Fpip.pypa.io\u002Fen\u002Fstable\u002Finstalling\u002F) >= 8.1.2\n- [setuptools](https:\u002F\u002Fsetuptools.readthedocs.io) >= 25.2.0\n- [wheel](http:\u002F\u002Fpythonwheels.com) >= 0.29.0\n- [numpy](http:\u002F\u002Fwww.numpy.org\u002F)\n- C++ 11 编译器 (如 gcc 4.8+ 或 clang)\n\n**macOS 额外要求：**\n- [Xcode command line tools](https:\u002F\u002Fdeveloper.apple.com\u002Flibrary\u002Fios\u002Ftechnotes\u002Ftn2339\u002F_index.html)\n\n## 安装步骤\n\n推荐使用二进制包进行安装。对于国内网络环境，建议配置 PyPI 镜像以提升下载速度。\n\n```bash\n# 使用清华镜像源安装 (推荐国内用户)\npip install -i https:\u002F\u002Fpypi.tuna.tsinghua.edu.cn\u002Fsimple nupic\n\n# 或使用官方源安装\npip install nupic\n```\n\n如需从源码安装（开发模式）：\n```bash\n# 安装本地源码\npip install .\n\n# 开发者模式安装 (-e 参数)\npip install -e .\n```\n\n> **注意**：若遇到权限问题，可尝试添加 `--user` 标志或在非系统环境下运行。Windows 用户可能需要管理员权限。\n\n## 基本使用与验证\n\n安装完成后，可通过运行单元测试来验证环境是否正确配置。\n\n```bash\n# 从仓库根目录执行测试\npy.test tests\u002Funit\n```\n\n### 后续开发\nNuPIC 的具体算法逻辑和 API 文档请参考官方文档：\n- 在线文档：\u003Chttp:\u002F\u002Fnupic.docs.numenta.org\u002F>\n- 社区论坛：\u003Chttps:\u002F\u002Fdiscourse.numenta.org\u002Fc\u002Fnupic>\n\n实际代码示例通常涉及导入 `nupic` 模块并构建 HTM 模型，请查阅上述文档获取详细用法。","某精密制造工厂的运维团队需要实时监控生产线关键电机的振动传感器数据，以提前发现潜在机械故障。\n\n### 没有 nupic-legacy 时\n- 传统固定阈值报警无法识别复杂时序模式，导致误报频发严重干扰正常生产节奏。\n- 主流深度学习模型依赖海量标注数据，冷启动阶段难以快速应用到新产线调试中。\n- 流式数据计算延迟较高，无法在资源受限的边缘端实时预测设备退化趋势。\n- 工况环境变化时需频繁重新训练模型，维护成本高昂且消耗大量算力资源。\n\n### 使用 nupic-legacy 后\n- nupic-legacy 基于 HTM 理论自动学习时空模式，大幅降低异常检测误报率并提升准确率。\n- 支持无监督连续学习，无需大量历史故障标签即可快速部署上线并适应新数据。\n- 内置高效流式处理机制，能在边缘设备毫秒级响应状态微小变化，实现即时预警。\n- 具备强自适应能力，随设备运行数据积累自动更新记忆库，显著减少人工调参与干预。\n\nnupic-legacy 凭借生物启发的算法架构，让工业设备在低算力边缘环境下实现了类脑智能的实时异常预警与长期趋势预测。","https:\u002F\u002Foss.gittoolsai.com\u002Fimages\u002Fnumenta_nupic-legacy_d393aaba.png","numenta","Numenta","https:\u002F\u002Foss.gittoolsai.com\u002Favatars\u002Fnumenta_1cf86ede.png","Biologically inspired machine intelligence",null,"http:\u002F\u002Fnumenta.com","https:\u002F\u002Fgithub.com\u002Fnumenta",[83,87,91,95,99,103,107,111],{"name":84,"color":85,"percentage":86},"Python","#3572A5",97.8,{"name":88,"color":89,"percentage":90},"Smarty","#f0c040",0.9,{"name":92,"color":93,"percentage":94},"Cap'n Proto","#c42727",0.5,{"name":96,"color":97,"percentage":98},"Shell","#89e051",0.4,{"name":100,"color":101,"percentage":102},"Jupyter Notebook","#DA5B0B",0.2,{"name":104,"color":105,"percentage":106},"PowerShell","#012456",0.1,{"name":108,"color":109,"percentage":110},"Dockerfile","#384d54",0,{"name":112,"color":113,"percentage":110},"Ruby","#701516",6357,1541,"2026-04-03T22:47:37","MIT",4,"Linux, macOS, Windows","未说明",{"notes":122,"python":123,"dependencies":124},"项目处于维护模式；必须使用 Python 2.7；macOS 需安装 Xcode 命令行工具；源码编译需 C++11 编译器；安装时可能需使用 --user 标志","2.7",[125,126,127,128,129],"pip>=8.1.2","setuptools>=25.2.0","wheel>=0.29.0","numpy","C++11 编译器 (gcc 4.8+ 或 clang)",[13,54],[132,133,134,135],"neocortex","artificial-intelligence","machine-intelligence","hierarchical-temporal-memory","2026-03-27T02:49:30.150509","2026-04-06T07:13:51.765524",[139,144,148,153,157,161],{"id":140,"question_zh":141,"answer_zh":142,"source_url":143},3185,"NuPIC 的可视化结果工具在哪里可以找到？","可视化功能的工作已移至独立的社区仓库。请访问 https:\u002F\u002Fgithub.com\u002Fnupic-community\u002Fnupic.visualizations 获取最新的可视化工具。","https:\u002F\u002Fgithub.com\u002Fnumenta\u002Fnupic-legacy\u002Fissues\u002F2658",{"id":145,"question_zh":146,"answer_zh":147,"source_url":143},3186,"如何获取或安装 NuPIC 的可视化脚本？","由于这是前端应用而非纯 Python 包，不建议使用 pip 安装。推荐直接使用 git clone 克隆仓库，或者下载渲染脚本以便离线使用。",{"id":149,"question_zh":150,"answer_zh":151,"source_url":152},3187,"运行示例时出现 `ImportError: No module named support.configuration` 错误怎么办？","这通常是由于环境变量配置不当导致的。请检查 `$NUPIC` 环境变量是否指向正确的源码目录，并确保没有创建嵌套文件夹（如 `\u002Fusers\u002Fusername\u002Fnupic\u002Fnupic`）。同时需要正确设置 `PYTHONPATH`。","https:\u002F\u002Fgithub.com\u002Fnumenta\u002Fnupic-legacy\u002Fissues\u002F2563",{"id":154,"question_zh":155,"answer_zh":156,"source_url":152},3188,"导入 `nupic.swarming` 模块时报错 `No module named swarming` 如何解决？","此错误同样源于环境配置问题。请确保 `PYTHONPATH` 包含了 NuPIC 的根目录，并且 `$NUPIC` 环境变量正确指向了包含 `nupic` 文件夹的路径。",{"id":158,"question_zh":159,"answer_zh":160,"source_url":152},3189,"运行 `~\u002Fexamples\u002Fswarm\u002Ftest_db.py` 脚本时数据库连接失败如何处理？","有用户反馈通过修改配置中的 IP 地址解决了此问题。尝试将脚本或配置中的 `'127.0.01'` 替换为 `'localhost'`，然后重新运行测试脚本。",{"id":162,"question_zh":163,"answer_zh":164,"source_url":165},3190,"在 Ubuntu 上安装后运行 Python 测试失败是什么原因？","测试失败通常与环境变量未正确设置有关。请确保已设置 `PYTHONPATH`, `NUPIC`, 和 `NTA` 环境变量。建议参考官方 Ubuntu 安装指南：https:\u002F\u002Fgithub.com\u002Fnumenta\u002Fnupic\u002Fwiki\u002FInstalling-NuPIC-on-Ubuntu。","https:\u002F\u002Fgithub.com\u002Fnumenta\u002Fnupic-legacy\u002Fissues\u002F769",[167,172,177,182,187,192,197,202,207,212,217,222,227,232,237,242,247,252,257,262],{"id":168,"version":169,"summary_zh":170,"released_at":171},102688,"1.0.5","* d81e128c6 NUP-2519: Update nupic.core version\r\n* 3371e4ef9 NUP-2519: Upgrade pycapnp to 0.6.3\r\n* 11a13c018 NUP-2518: Remove obsolete region initialization parameters from custom region example\r\n* 67724debf \"pip install --use-wheel\" was deprecated. See https:\u002F\u002Fpip.pypa.io\u002Fen\u002Fstable\u002Fnews\u002F#deprecations-and-removals\r\n* 08daff4a3 Fix softmax overflow\r\n","2018-06-01T15:12:12",{"id":173,"version":174,"summary_zh":175,"released_at":176},102689,"1.0.4","* 682fd2e66c6d45912cd9c518106740143eff9fd9 :  NUP-2506: Add test to all Serializable subclasses and fix related issues  (#3826)\r\n* a7ab556a64e57064a8febbcf2ee341a1b1ff18ae : NUP-2506: fix traversal limit (#3823)\r\n* 54e1ffead7a8cedd9a2dab08bb02c7e5e3536bf3 : Added holidays parameter to date encoder (#3822)\r\n* 9bb7705ebd428e73f6efc180e77f7f127ce67c4d : version lock 'sphinx-autobuild' dependency 'tornado' (#3815)\r\n* 13c02de82c6bc52afc364fd1fdce88c5fa1aa92c : Update some legacy code examples. (#3814)\r\n* 33052e10dbe1030929223fc7e54d2d7c8b8a1ced : Fix metric spec schema bug (#3812)\r\n* 40e216915a172901b5dbe34932543fae68aa3631 : Fix lack of logging in run_swarm.py (#3809)\r\n* d8c740486198a6764b18c73bffe6005988435ac7 : NUP-2487: Update category prediction example\r\n* 38e40266a7ad2744fce904a02faa383676950e1e : Issue #1380: Update SP parameter validation test checking array dimensions\r\n* e470860e962db70a2b16c82d1647f10b3e985c42 : numenta\u002Fnupic.core#1380: Fix SP tests with correct dtype values\r\n* 38c9c7e1d7b161d9a4b2378cdae3652af704a481 : Add example for infer as well\r\n* 94e5f62e669dcbd55268c07b6aa30f391135ab4f : Include example to make isSparse parameter easier to understand\r\n* ffd1457037a52cb63a7be0cf004e886f3909f505 : Update KNN classifier documentation to make the input pattern requirements clear in both learn and infer\r\n* 5ecae91017c5f4f68c944a3f5b5d79d1276e2c59 : Changed distribution keyword, casted some attributes to float, removed setting list (#3784)\r\n* 41e5a6aefc649b08dbe948ba3e1db46f5aaaa603 : Issue #3783: Fixes test to compute pass the probation period (#3786)\r\n* 7e5f587ecd039520a53a4aa4608eb7ce577654f1 : Updating to XCode 8.3\r\n* 1aea72abde4457878a16288d6786ffb088f69164 : Update name of nyc_taxi.csv to nycTaxi.csv (#3776)\r\n","2018-04-12T18:51:05",{"id":178,"version":179,"summary_zh":180,"released_at":181},102690,"1.0.3","Release 1.0.3.\r\n\r\n- Updated incorrect name for anomaly likelihood region. (#3770)\r\n\r\n","2017-09-13T18:24:07",{"id":183,"version":184,"summary_zh":185,"released_at":186},102691,"1.0.2","Release 1.0.2\r\n\r\n- Fixed BacktrackingTM serialization error (#3765)","2017-08-22T20:55:33",{"id":188,"version":189,"summary_zh":190,"released_at":191},102692,"1.0.1","- Fixed a bug in record sensor that prevented the usage of CoordinateEncoder. (#3754)","2017-08-08T20:01:55",{"id":193,"version":194,"summary_zh":195,"released_at":196},102693,"1.0.0","- Improved exception handling in \u002Fswarm\u002Ftest_db.py (#3738)\r\n- DEVOPS-362 Remove unnecessary install script\r\n- DEVOPS-362 test removing dependency on install script entirely\r\n- DEVOPS-362 update install script\r\n- DEVOPS-362 Add initial version of missing install script\r\n- Added serialization guide to API docs. (#3737)\r\n- Put conditional around capnp for Windows\r\n- Complete new serialization in SpatialPooler\r\n- Catch nupic.core reference\r\n- NUP-2342: consolidate read\u002Fwrite into a single context\r\n- NUP-2342: Update examples to use capnp serialization\r\n- NUP-2341: Use capnp serialization for SDRClassifierDiff\r\n- NUP-2351: Remove TODOs from HTM Prediction Model test and fix  bugs exposed by this test\r\n- NUP-2351: Add serialization to KNNAnomalyClassifierRegion\r\n- NUP-2351: Fix KNNClassifier serialization\r\n- NUP-2349 Implemented testCapnpWriteRead test for PreviousValueModel OPF class. Implemented PreviousValueModel.getProtoType. Return instance from PreviousValueModel.read.\r\n- NUP-2349 Implemented capnp serialization of PreviousValueModel\r\n- Put capnp import checks in place for Windows\r\n- Add serialization tests for TMRegion\r\n- NUP-2463 Serialize inferenceArgs, learningEnabled, and inferenceEnabled in opf Model.\r\n- Add support for different TM types in TMRegion serialization\r\n- Added Serializable to API docs, and inheritence links\r\n- Fixed Next ID value in comment in model.capnp\r\n- NUP-2464 Integrated ModelProto support into opf TwoGramModel.\r\n- Fixed input to SP in docs algo example (#3708)\r\n- NUP-2464 Serialize numPredictions and inferenceType via ModelProto member of HTMPredictionModelProto.\r\n- Added Serializable to all classes with a capnp write function (#3710)\r\n- Safe import of capnp for moving average proto\r\n- getSchema returns prototype\r\n- Remove unused \\_readArray\r\n- Rely on pycapnp\u002Fnumpy native conversions in write\u002Fread\r\n- Add capnp conditionals for Windows\r\n- NUP-2351: Use dict directly instead of creating capnp message\r\n- Fixed Serializable extensions\r\n- Fix CPP breakages from changes\r\n- NUP-2351: Add capnp serialization to KNNClassifierRegion\r\n- Fix everything up to get serialization tests working with capnp serialization for BacktrackingTM\r\n- Added getSchema to MovingAverage\r\n- Added Serializable to all classes with a capnp write function\r\n- Finished up first pass implementation of BacktrackingTM serialization\r\n- NUP-2350: capnp serialization for TwoGramModel\r\n- NUP-2449 Completed implementation of HTMPredictionModel serialization tests.\r\n- NUP-2463 Implemented test (disabled) to demonstrate the bug \"Predicted field and \\_\\_inferenceEnabled are not serialized by HTMPredictionModel.write\"\r\n- OPF Guide cleanup and link fixes (#3700)\r\n- NUP-2355 Add new serialization to TestRegion\r\n- remove SVMClassifierNode (#3697)\r\n- handle scalar values in the sdr classifier region\r\n- NUP-2346: Add serialization to knn\\_classifier\r\n- NUP-2458 Fixed and enabled SDRClassifierTest.testWriteReadNoComputeBeforeSerializing\r\n- NUP-2458 Implemented testWriteReadNoComputeBeforeSerializing in sdr\\_classifier\\_test.py that reproduces the \"deque index out of bounds\", but disabled the test, since it fails in a different way after the fix, most likely unrelated to the fix, which needs to be debugged\r\n- NUP-2398 Refactor test comparing different configurations\r\n- NUP-2458 Prevent index out of bounds when saving `patternNZHistory` after fewer than \\_maxSteps input records have been processed.\r\n- NUP-2458 Moved HTMPredictionModel serialization test to integration\u002Fopf\r\n- NUP-2449 Implement simple serialization\u002Fdeserialzation tests. This exposed a number of problems that need to be fixed before we can make further progress.\r\n- update sdr classifier doc","2017-07-07T20:04:14",{"id":198,"version":199,"summary_zh":200,"released_at":201},102694,"0.8.0","* Document ExperimentDescriptionAPI (#3679)\r\n* Update nupic.math API docs (#3677)\r\n* SP docs cleanup (#3671)\r\n* Allow multiple classifications for each record to SDRClassifier (#3669)\r\n* Updated BacktrackingTMCPP compute parameter name (#3667)\r\n* Fix HTMPredictionModel prediction using SDRClassifier (#3665)\r\n* Remove CLAClassifier (#3665)\r\n* Add capnp serialization to TMRegion (#3657)","2017-06-08T20:31:16",{"id":203,"version":204,"summary_zh":205,"released_at":206},102695,"0.7.0","## 0.7.0\r\n\r\n**WARNING**: This release contains breaking changes described in\r\nhttps:\u002F\u002Fdiscourse.numenta.org\u002Ft\u002Fwarning-0-7-0-breaking-changes\u002F2200\r\n\r\n* Stop calling the backtracking_tm tests \"tm tests\" (#3650)\r\n* Update hierarchy demo to fix regression\r\n* Clean up BacktrackingTM's public API (#3645)\r\n* Make region file names snake_case (part 7) (#3640)\r\n* Removed references to obsolete tm_py_fast shim (#3639)\r\n* Updated OPF Metric API docs (#3638)\r\n* updated __init__.py to include missing encoders (#3487)\r\n* Fixed anomaly likelihood doc problems. (#3629)\r\n* Updates swarming, some region code to snake_case (part 6) (#3627)\r\n* Fixed OPF util helpers module names. (#3625)\r\n* Complete RST docs for nupic.support (#3624)\r\n* Deleted nupic.support.features* (unused) (#3622)\r\n* Removed nupic.support.exceptions (unused) (#3620)\r\n* Proper snake_case for nupic.support (part 5) (#3618)\r\n* Snake case nupic.encoders (part 4) (#3614)\r\n* Moved opf_helpers module to helpers (#3610)\r\n* Removes unused code from nupic.support (#3616)\r\n* Applying snake_case module name standards (PART 3) (#3611)\r\n* Fixed support initLogging docstring params\r\n* Finished OPF utils and env docs\r\n* Documented OPF Basic Env\r\n* Documented OPF ENv\r\n* Documenting OPF Task Driver\r\n* Documenting OPF experiment runner\r\n* Removed OPF utils PredictionElement (#3604)\r\n* Partial doc of experiment description api\r\n* NUP-2429 Add .gitignore with first_order_0.csv to prevent accedental commits of this generated file.\r\n* Documented cluster_params canned model config\r\n* Documented OPF model exceptions\r\n* Finished doccing opf_utils\r\n* Documenting OPF utils\r\n* Removed predictedField from HTMPredictionModel constructor (#3600)\r\n* NUP-2420 Renamed tm_shim.py to BacktrackingTM_shim.py\r\n* Removes inputRef \u002F bookmark params from appendRecord (#3597)\r\n* Documented nupic.data (#3593)\r\n* OPF Model docstrings (#3589)\r\n* Remove obsolete nupic.research.bindings check\r\n* Removed unimplemented abstract methods (#3596)\r\n* Removed WeatherJoiner code from old example (#3595)\r\n* Updated snakecase opf_utils in RST docs (#3585)\r\n* Renamed tm_ccp test so it runs\r\n* Moved research tm_cpp_test.py back into nupic.research\r\n* Removed base.Encoder.formatBits() (#3582)\r\n* Replace dump() with define __str__ in Encoders. Issue #1518 (#3559)\r\n* Complete encoder docstrings (#3579)\r\n* Removed nupic.research, moved contents to nupic.algorithms\r\n* move zip logic into 'build_script'\r\n* Add support for artifacts deployed to S3 named according to sha\r\n* Snake case module names PART 2 (#3561)\r\n* Remove old examples Part 2 (#3562)\r\n* NUP-2401: Check for prediction results discrepancies (#3558)\r\n* NUP-2397: rename TP* to TM* (#3555)\r\n* NUP-2405: quick-start guide for the Network API (#3557)\r\n* Snake case module names PART 1 (#3550)\r\n* NUP-2394: network API code example (#3520)\r\n* Remove old examples Part 1 (#3551)\r\n* Docs: InferenceShifter,ModelResult,SensorInput,InferenceType (#3549)\r\n* CLAModel name changed to HTMPredictionModel (#3516)\r\n* Updating FileRecordStream docstrings (#3545)\r\n* Fieldmeta docstrings (#3541)\r\n* Update KNNClassifier docstrings (#3535)\r\n* SDRClassifier docs, default config docs\r\n* Updates anomaly docstrings (#3537)\r\n* [NUP-2399] Added style guides to new guide (#3528)\r\n* NUP-2396 Allow SensorRegion to pass actValue and bucketIdx to SDRClassifierRegion\r\n* Added anomaly detection guide (#3521)\r\n* NUP-2389 Upgrade nupic.bindings dependency to 0.6.1 which has the requisite changes.\r\n* name change tpParams\u002FtmEnable => tmParams\u002FtmEnable (#3514)\r\n* NUP-2391: packages to document & progress tracking (#3517)\r\n* Quick Start\r\n* NUP-2389 Remove calls to Region::purgeInputLinkBufferHeads. Since we only support delay=0 in CLA models, we no longer need `purgeInputLinkBufferHeads`, because the new Link::compute logic in nupic.core now performs direct copy from src to dest for links with delay of 0.\r\n* Disable flatline hack in anomaly likelihood\r\n","2017-06-02T19:47:30",{"id":208,"version":209,"summary_zh":210,"released_at":211},102696,"0.6.0","## 0.6.0\r\n\r\n* Touch init even if model params dir exists\r\n* Auto-add __init__.py when model parms created\r\n* Shift code from otherwise unused `nupic.engine.common_networks` to example where it's used.  Includes bugfix renaming `rawAnomalyScore` to `anomalyScore`\r\n* Explicitly import and use `engine_internal` in lieu of `engine` to avoid confusion, create `nupic.engine.OS` and `nupic.engine.Timer` by assignment rather than subclass\r\n* Change SparsePassThroughEncoder dtype error to ValueError\r\n* Fix for an unrelated change that resulted in numpy arrays being used in cpp implementation\r\n* Give better message for bad dtype to SparsePassThroughEncoder\r\n* Add test for passing float values for radius\r\n* Adds api docs for coordinate encoders\r\n* Cleanup CoordinateEncoder\r\n* Remove svm, cells4 tests that are moved to nupic.core.\r\n* Added missing anomaly stuff, fixed requirements\r\n* Moved sphinx deps out of requirements.txt\r\n* Fix hotgym_regression_test.py to make it work with nupic.core PR 1236.\r\n* Skip test when capnp is not available, such as windows as well as address feedback from Scott\r\n* Serialization base python class analagous to nupic.core Serializable c++ class\r\n* Adds a demo Jupyter notebook, useful for demonstrating usage of visualization framework and as an entrypoint for tinkering with different network topologies\r\n* Speed up SpatialPooler read method.\r\n* Rename normalProbability to tailProbability.\r\n* Use IterableCollection from engine_internal\r\n* Call Region.purgeInputLinkBufferHeads after compute() calls in CLAModel to integrate with the new delayed link implementation from nupic.core.\r\n* rename maxBoost to boostStrength in hotgym example\r\n* Disable backward compatibility serialization test\r\n* remove minPctActiveDutyCycle parameter form SP compatability test\r\n* update expected result in hotgym, result change due to different rounding rules in boosting\r\n* eliminate minPctActiveDutyCycle from spatial pooler\r\n* Rename maxBoost to BoostStrength\r\n* Stop changing the overlaps to do tie-breaking\r\n* Stop trying to get identical boost factors between py and cpp\r\n* set maxBoost in expdescriptionapi\r\n* update sp_overlap_test to use global inhibition\r\n* slight simplification of boostFactor calculation\r\n* Implement update boost factors local and global\r\n* Avoid floating point differences with C++ SpatialPooler\r\n* run C++ SP in spatial_pooler_boost_tests\r\n* update spatial pooler boost test\r\n* update boosting rules for spatial pooler\r\n* fix bug in setPotential\r\n* modified SP boosting rule","2017-03-30T18:10:12",{"id":213,"version":214,"summary_zh":215,"released_at":216},102697,"0.5.7","- Remove tests moved to nupic.core and update version to latest bindings release.\n- Update hello_tm.py\n- Removed linux and gcc from Travis build matrix\n- Makes `anomaly_likelihood.py` compliant to Python3\n- Update env vars and paths to simplify the AV configuration and installation.\n- Cleanup references to nupic.bindings and old CI code for manually fetching nupic.bindings since it should be found on PyPI without doing anything special.\n","2016-11-28T23:01:58",{"id":218,"version":219,"summary_zh":220,"released_at":221},102698,"0.5.6","- Since manylinux nupic.bindings wheel 0.4.10 has now been released to PyPi, we no longer need to install nupic.bindings from S3.\n- fix logic in _getColumnNeighborhood\n- Bugfix in flatIdx reuse after a segment is destroyed\n- Change private _burstColumn class method signature to accept a cellsForColumn argument in lieu of a cellsPerColumn argument.  Move the calculation that otherwise depends on cellsPerColumn into the instance method.\n- TM: Support extensibility by using traditional methods\n- Update expected error for topology changes\n- Update expected hotgym result for topology changes\n- Adds RELEASE.md with documentation for releasing NuPIC.\n- Match nupic.core's SP neighborhood ordering.\n- Update inhibition comments and docstrings.\n- Introduce mechanism by which already-installed pre-release versions of nupic.bindings are ignored during installation\n- Assign self.connections value from self.connectionsFactory() rather than direct usage of Connections constructor. Allows better extensibility should the user want to change some aspect of the creation of the connections instance in a subclass\n- Removed obsolete directory src\u002Fnupic\u002Fbindings\u002F\n- Remove the notion of \"destroyed\" Segments \u002F Synapses\n- Enable proper subclassing by converting staticmethods that referenced `TemporalMemory` to classmethods that reference their class.\n- Fixup TemporalMemory.write() to handle columnDimensions as tuples.\n- Initialize columnDimensions as a tuple in test to reflect common convention.  This forces the TemporalMemoryTest.testWriteRead test to fail in its current state.\n- Store \"numActivePotentialSynapses\". No more \"SegmentOverlap\".\n- Add a lot more scenarios to the TM perf benchmark\n- Moved audiostream example to htm-community\n- Safer \"addToWinners\" value. Play nicely with surgical boosting.\n- Bugfix: With no stimulus threshold, still break ties when overlaps=0\n- Clean up trailing whitespace and tabs\n- Properly apply the stimulus threshold\n- Add test for new \"learn on predicted segments\" behavior\n- Split compute into activateCells and activateDendrites\n- Grow synapses in predicted columns, not just bursting columns\n- Removed bundled get-pip.py and instead fetch version copy from S3\n- Removed .nupic_modules and now rely on versioned release of nupic.bindings on PyPI\n- averagingWindow size updated to improve HTM scores for RES-296\n- Build system updates for Bamboo (Linux), Travis (OS X), and AppVeyor (Windows)\n- Added nyc taxi example for anomaly detection\n","2016-10-20T17:10:36",{"id":223,"version":224,"summary_zh":225,"released_at":226},102699,"0.5.5","- Renamed a misclassed class name from ConnectionsTest to GroupByTest\n- not _ is => is not and fixes groupby comment and passes integration tests\n- overhaul to groupby, now 10% faster than current implementation\n- NUP-2299 Install specific versions of pip, setuptools, and wheel.\n- NUP-2299 Added platform-conditional dependency on pycapnp==0.5.8 using PEP-508.\n- lazy group_by and changes to GroupByGenerator\n- perf improvement to segment comparison in compute activity\n- 100 % increase in spped\n- small perf changes\n- demonstrate that compatability test works with predictedSegmentDec not 0.0\n- fixes subtle bug in numSegments that caused integration tests to fail\n- fixes bug where minIdx could be passed as a float rather than an int\n- skip serialization test if capnp is not installed\n- lints and updates comments in group_by.py and group_by_tests.py\n- gets same results as c++ temporal memory after group_by changes\n- ports group_by tests and they pass\n- adds groupByN utility function for use in TM\n- all connections tests written and passing, moved some stuff around and added missing function to connections\n- started porting new connections tests and minor changes to connections.py\n- improves permanence >= testing in computeActivity\n- confirmed python implementation is same as cpp version. Needs better perf now\n- adds back AnomalyRegion and Anomaly class in anomaly.py and related tests\n- fixes bug in growSynapses, almost exactly the same\n- Updated core SHA and default SDR classifier implementation\n- Updated SDRClassifier factory and region to handle cpp\n- changed input name from value to metricValue\n- updates variables names in anomaly_likelihood.py and AnomalyLikelihoodRegion\n- adds new connections methods\n- create new methods for creating\u002Fdestroying synapses\u002Fsegments\n- continues change of connections datastructures\n- move raw anomaly calculation back to nupic.algorithms.anomaly\n- Finished swarming\u002Fhypersearch separation\n- Moved base hypersearch classes to hypersearch\n- Moved experimentutils to nupic.swarming\n- Updated SDR classifier internals\n- calculate raw anomly score in KNNAnomalyClassifier\n- removes anomaly.py dependency in network_api_demo.py\n- changes how TPRegion computes prevPredictdColumns and updates clamodel\n- Install pip from local copy, other simplifications\n- Fixup PYTHONPATH to properly include previously-defined PYTHONPATH\n- adds pseudocode to core functions\n- continues implementation of AnomalyLikelihoodRegion\n- Limit tests to unit after ovverriding pytest args on cli\n- DEVOPS-85 OS X build infrastructure for Bamboo CI environment\n- replaces segmentCMP with lambda and updates docstrings\n- uses arrays instead of dicts in computeActivity\n- Corrections to examples in tm_high_order.py\n- incorporates binary search into the algorithm where applicable\n- remove outdated nab unit tests\n- use Q function\n- Corrections to examples in tm_high_order.py\n- change to column generator\n- Added tm_high_order.py to show examples of the temporal memory.\n- Fixed conversion bug in SDRClassifier serialization\n- Fixed patternNZ proto writing.\n- Slight fix for pattern history handling in sdr classifier\n- Small fix on SDR classifier\n- Better fix for #3172, using the initialize() function and checking if _sdrClassifier is set\n- Updated learning rate for SDR classifier + slight changes to the error ranges in OPF test\n- Updated hotgym test with actual value and implemented first fix for OPF test\n- Updated tests and examples with SDR classifier\n- Finished updating examples with SDR classifier.\n- Updated hotgym and general anomaly examples with SDR classifier.\n- Updates pycapnp to 0.5.8\n- test_db-fixes avoids printing user password in plaintext\n- test_db-fixes updates database and table name\n- Corrections made to the spatial pooler tutorial.\n- changes maxBoost default value to 1.0\n- fixes connection tests and prints config file used in test_db.py\n- Moved back overlap accesors test for spatial_pooler from API tests to unit tests.\n- Added tutorial script for the spatial pooler. Modified README file accordingly.\n- Moved the unit test for SP overlap accesors to API tests.\n","2016-08-17T20:08:43",{"id":228,"version":229,"summary_zh":230,"released_at":231},102700,"0.5.4","- Added overlap accessors to spatial_pooler.py plus unit tests. (Code style corrected)\n- Updated VERSION in Spatial Pooler and added backward compatibility in setstate()\n- Added members overlaps and boostedOverlaps to SpatialPooler class.\n- Addition of overlaps and boostedOverlaps members to SpatialPooler class plus unit tests.\n- Added docs for return type in RDSE internal func.\n- tm_cpp with tuned parameters\n- RES-215 Changes to add params for new TM subclass for NAB\n- Remove main function from SDRClassifierRegion\n- remove unused methods from SDRClassifierRegion\n- Add simple end-to-end integration test for SDRClassifierRegion\n- use string split instead of eval to parse strings\n- correct inconsistent error msg in sdr_classifier_factory.py\n- Fix readWrite test of SDR classifier\n- Add SDRClassifier Region to pyRegions\n- Initial implementation of SDRClassifier Region\n- implement SDR classifier factory\n- Add capnp proto for SDR classifier region\n- Add default value for SDR classifier implementation in nupic-default.xml\n","2016-06-02T19:09:30",{"id":233,"version":234,"summary_zh":235,"released_at":236},102701,"0.5.3","- Default DATETIME columns to NULL in ClientJobsDAO for compatibility across mysql versions. As of mysql 5.7.8, values of 0 are not allowed for DATETIME columns, and CURRENT_TIMESTAMP is semantically inappropriate for those columns.\n- Suppress this optional dependency on matplotlib without logging, because python logging implicitly adds the StreamHandler to root logger when calling `logging.debug`, etc., which may undermine an application's logging configuration\n- Bugfix: Write the 'actualValues' to the output, don't reassign the output\n- Fixed Username Regex in ClientJobsDAO\n- cleaned up region a bit to make it compliant with numenta's coding guidelines.\n","2016-05-16T21:44:50",{"id":238,"version":239,"summary_zh":240,"released_at":241},102702,"0.5.2","- Fixe to GCE to return the right number of scalars when altitude is missing.\n","2016-04-15T17:48:16",{"id":243,"version":244,"summary_zh":245,"released_at":246},102703,"0.5.1","- Improves SDR classifier and tests\n- Modify the continuous online learning test\n- Add 3 tests on multiple item prediction\n- Fix test_pFormatArray\n- Implement SDR classifier in NuPIC\n- Make the 'arrayTypes' list more informative\n- Add getParameter\u002FsetParameter support for Bool and BoolArray\n- Improved anomaly params (from NAB)\n- Added minSparsity option\n- Get the encoder's outputWidth via parameter\n- Use nupic.core encoders from nupic via the Network API\n- Fix bugs and inconsistencies in the custom region demo\n- Adds BINDINGS_VERSION envvar to wheel filename (for iterative builds)\n","2016-03-28T19:48:38",{"id":248,"version":249,"summary_zh":250,"released_at":251},102704,"0.5.0","- Removes references to FastTemporalMemory.\n- Lower TM epsilon threshold for compatibility.\n- Add documentation for the Monitor Mixins\n- Removed FastTemporalMemory from nupic\n- Update temporal memory compatibility test to use C++ TM.\n- Sort segments before iterating for compatibility with C++\n- Sort unpredictedActiveColumns before iterating for compatibility with C++\n","2016-02-15T20:18:27",{"id":253,"version":254,"summary_zh":255,"released_at":256},102705,"0.4.5","- This release is just to sync with nupic.bindings 0.3.1.\n","2016-01-27T04:13:26",{"id":258,"version":259,"summary_zh":260,"released_at":261},102706,"0.4.3","- Updating to proper core sha\n","2016-01-26T23:07:39",{"id":263,"version":264,"summary_zh":265,"released_at":266},102707,"0.4.2","- Using official release version of bindings for nupic release.\n","2016-01-25T18:27:42"]