{"base_url":"https://www.spiritysdx.top/","description":"个人技术记录、工程排障和系统实践日志","generated_at":"2026-06-17T12:24:10+08:00","language":"zh-CN","llms":"https://www.spiritysdx.top/llms.txt","mcp":"https://www.spiritysdx.top/.well-known/mcp.json","posts":[{"categories":["电脑技巧","Kubernetes"],"date":"2025-05-23T16:16:53+08:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20250523-ray-ascend-npu/index.md","section":"posts","summary":"这篇记录 Ray 在 Kubernetes 中对接昇腾 NPU 的流程，包括 Ray 镜像制作、KubeRay Operator 安装、RayCluster 编排和资源验证。环境基于 aarch64，镜像中需要带入宿主机昇腾驱动运行所需文件。 ","tags":["ray","kuberay","kubernetes","ascend","npu","docker"],"title":"Ray 支持昇腾 NPU 的 KubeRay 对接记录","url":"https://www.spiritysdx.top/20250523-ray-ascend-npu/"},{"categories":["电脑技巧","Linux"],"date":"2025-05-23T14:19:06+08:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20250523-openeuler-system-adaptation/index.md","section":"posts","summary":"这篇记录 openEuler 22.03 LTS SP4 arm64 环境下制作离线依赖包的过程。目标是对齐已有 amd64 离线源，补齐 arm64 目录中缺失的 Docker、NFS 和基础工具包，并在纯净环境中验证安装冲突。 ","tags":["linux","openeuler","arm64","rpm","dnf","docker"],"title":"openEuler 22.03 arm64 离线包适配记录","url":"https://www.spiritysdx.top/20250523-openeuler-system-adaptation/"},{"categories":["电脑技巧","Docker"],"date":"2025-05-12T14:22:46+08:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20250425/index.md","section":"posts","summary":"这篇记录在线 IDE 或浏览器可访问应用的通用容器镜像打包流程。重点是基础镜像选择、软件源、中文环境、GPU/NPU 驱动、图形界面、SSH、端口转发和镜像导入导出这些容易踩坑的环节。 ","tags":["docker","ide","dockerfile","jupyter","jetbrains","vnc","novnc","kasm","gpu"],"title":"在线 IDE 的通用容器镜像打包流程","url":"https://www.spiritysdx.top/20250425/"},{"categories":["电脑技巧","Linux"],"date":"2025-05-12T14:11:01+08:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20250512-kylin-v10-local-repo/index.md","section":"posts","summary":"麒麟 V10 安装时如果没有正确指定本地镜像源，安装界面可能报错： 1 设置基础软件仓库时出错 这个问题需要在进入图形安装界面之前处理，也就是在启动引导界面修改内核参数。 操作步骤 在安装引导界面，将光标移动到： ","tags":["linux","kylin","repo","installation"],"title":"麒麟 V10 安装时配置本地源","url":"https://www.spiritysdx.top/20250512-kylin-v10-local-repo/"},{"categories":["电脑技巧","Kubernetes"],"date":"2025-05-12T12:03:20+08:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20250512-bmp-troubleshooting/index.md","section":"posts","summary":"这篇整理 BKE/BMP 部署和运行阶段的常用排障流程。核心原则是先确认当前处于部署前、部署中还是部署后的集群运行阶段，再选择对应的 bke、kubectl、docker 或系统服务排查路径。 ","tags":["kubernetes","k3s","docker","bke","bmp","troubleshooting"],"title":"BKE 与 BMP 部署排障流程","url":"https://www.spiritysdx.top/20250512-bmp-troubleshooting/"},{"categories":["电脑技巧","Docker"],"date":"2025-05-12T11:15:37+08:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20250512-containerd-http-registry/index.md","section":"posts","summary":"这篇记录 containerd 拉取 HTTP 镜像仓库时的配置方式。常见误区是只在 /etc/containerd/config.toml 中配置 insecure = true，但这个配置主要作用于 CRI 插件；使用 nerdctl、ctr 或 Kubernetes 拉取时，仍可能继续尝试 HTTPS。 ","tags":["containerd","docker","kubernetes","registry","http"],"title":"containerd 允许 HTTP 镜像仓库拉取","url":"https://www.spiritysdx.top/20250512-containerd-http-registry/"},{"categories":["电脑技巧","Linux"],"date":"2025-05-12T10:54:21+08:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20250512-redhat-system-adaptation/index.md","section":"posts","summary":"这篇记录整理 Red Hat 8 amd64 环境下为 BKE 部署准备离线依赖包的流程。核心思路是先在可联网环境中补齐 RPM 及其依赖，再把离线包复制到纯净测试机中验证安装，最后进入 allinone 部署。 ","tags":["linux","redhat","centos","rpm","dnf","bke"],"title":"Red Hat 8 amd64 离线依赖适配记录","url":"https://www.spiritysdx.top/20250512-redhat-system-adaptation/"},{"categories":["Golang"],"date":"2025-04-15T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20250415/index.md","section":"posts","summary":"确认当前k8s版本以及对应的kubevirt版本 1 kubectl version 显示 1 2 Client Version: version.Info{Major:\u0026#34;1\u0026#34;, Minor:\u0026#34;23\u0026#34;, GitVersion:\u0026#34;v1.23.17\u0026#34;, GitCommit:\u0026#34;953be8927218ec8067e1af2641e540238ffd7576\u0026#34;, GitTreeState:\u0026#34;clean\u0026#34;, BuildDate:\u0026#34;2023-02-22T13:34:27Z\u0026#34;, GoVersion:\u0026#34;go1.19.6\u0026#34;, Compiler:\u0026#34;gc\u0026#34;, …","tags":["golang","kubevirt","kubernetes","kvm","windows","k8s"],"title":"kubevirt初体验","url":"https://www.spiritysdx.top/20250415/"},{"categories":["电脑技巧","docker"],"date":"2025-04-05T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20250405/index.md","section":"posts","summary":"修改dockur制作Windows镜像 https://github.com/dockur/windows 拉取测试的版本: v4.33 原始仓库的镜像默认只是一个ISO下载器和网络自动设置器，本质上不包含Windows镜像，容器启动后镜像会将系统安装到挂载出的盘中。 ","tags":["docker","dockur","kvm","windows"],"title":"魔改dockur制作可迁移的Windows镜像(单文件)","url":"https://www.spiritysdx.top/20250405/"},{"categories":["电脑技巧","docker"],"date":"2025-03-15T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20250315/index.md","section":"posts","summary":"使用dockur的原始镜像制作Windows镜像(双文件) https://github.com/dockur/windows 原始仓库的镜像默认只是一个ISO下载器和网络自动设置器，本质上不包含Windows镜像，容器启动后镜像默认会将系统安装到挂载出的盘中。 本页说明最终将保存两个文件，一个是挂载盘的压缩文件，一个是容器导出的tar包。 ","tags":["docker","dockur","kvm","windows"],"title":"通过dockur制作可迁移的Windows镜像(双文件)","url":"https://www.spiritysdx.top/20250315/"},{"categories":["电脑技巧","Golang"],"date":"2025-03-15T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20250325/index.md","section":"posts","summary":"通过ProxmoxVE制作kubevirt可用的WIN镜像 https://github.com/ILLKX/Windows-VirtIO 下载带virtio的虚拟机镜像 然后借鉴 https://www.spiritlhl.net/guide/pve/pve_windows.html 开设虚拟机，到图形化安装后即可，不要配置网络 在虚拟机内需要设置一个bat脚本，设置系统启动后执行，脚本的内容是 ","tags":["golang","kubevirt","kubernetes","proxmox","proxmoxve","kvm","windows","k8s"],"title":"通过ProxmoxVE制作kubevirt可用的Windows镜像","url":"https://www.spiritysdx.top/20250325/"},{"categories":["电脑技巧","docker"],"date":"2025-03-05T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20250305/index.md","section":"posts","summary":"仓库 https://github.com/oneclickvirt/dockerfile-templates 适配境内环境，搭建带浏览器web查看NOVNC的在线IDE 对应开发环境支持 AMD64 ARM64 对应GPU/NPU的驱动支持 对应平台的AI插件支持 由于版权问题和镜像大小问题，tar包我就不传了 ","tags":["docker","ide","dockerfile","jupyter","webstorm","idea","pycharm","eclipse","vscode","jetbrains","vnc","kasm"],"title":"基于docker的在线IDE制作(支持浏览器直接访问)","url":"https://www.spiritysdx.top/20250305/"},{"categories":["Golang"],"date":"2025-02-25T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20250225/index.md","section":"posts","summary":"一般流程 1 free -m 需要确保无SWAP，否则kubelet起不来 然后需要 1 docker ps -a | grep etcd 看平面容器起来了没有，没有的话就得看容器日志排查问题 然后 1 kubectl get pods -A 看核心的namespace的pod有没有起来，有没有ready，有问题的pod名字就describe一下 下面所有示例都用kube-system作为查询的namespace，实际看你要查什么服务对应的namespace ","tags":["golang","kubernetes","k8s"],"title":"k8s日常问题排障","url":"https://www.spiritysdx.top/20250225/"},{"categories":["电脑技巧","Golang"],"date":"2024-07-14T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20240714/index.md","section":"posts","summary":"问题 目前新的go项目默认都是go module模式，由于我需要使用私有仓库的package，清除了mod的缓存后Goland就识别不到我后续在命令行下执行 ","tags":["golang","goland"],"title":"jetbrains家的goland项目可用但老爆红","url":"https://www.spiritysdx.top/20240714/"},{"categories":["电脑技巧"],"date":"2024-07-13T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20240713/index.md","section":"posts","summary":"前言 由于Github在国内访问速度较慢，所以选择在本地服务器上自建Gitea，使用国内服务器内穿端口后访问速度较快，并且可以自己管理代码，更安全。 ","tags":["linux","ubuntu","gitea","ssh"],"title":"机房本地服务器自建Gitea并使用","url":"https://www.spiritysdx.top/20240713/"},{"categories":["电脑技巧"],"date":"2024-07-12T00:00:00+00:00","lastmod":"2025-06-28T18:47:09+08:00","markdown_url":"https://www.spiritysdx.top/20240712/index.md","section":"posts","summary":"问题 如题目所说，这里贴个报错 1 2 3 nvidia-smi Failed to initialize NVML: Driver/library version mismatch NVML library version: 535.183 原先的版本是 1 NVIDIA-SMI 535.171.04 Driver Version: 535.171.04 CUDA Version: 12.2 修复方案 下载官方驱动 打开官方网站：https://www.nvidia.cn/Download/Find.aspx?lang=cn 按照你的显卡版本进行选择， …","tags":["linux","ubuntu","gpu","conda"],"title":"nvidia-smi被自动升级无法与GPU通信了怎么办","url":"https://www.spiritysdx.top/20240712/"},{"categories":["机器学习","Python"],"date":"2024-06-07T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20240607/index.md","section":"posts","summary":"conda默认的python环境版本过高，需要降级默认的Python环境 假设原始环境是Python12的环境 直接使用 1 conda install python==3.11 类似上述方法指定python版本下载，注意这个命令要在conda的虚拟环境中执行才会替换当前python版本。 ","tags":["linux","ubuntu","gpu","conda"],"title":"conda使用GPU时的一些陷阱","url":"https://www.spiritysdx.top/20240607/"},{"categories":["电脑技巧"],"date":"2024-05-14T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20240515/index.md","section":"posts","summary":"前言 由于机房的服务器(下称宿主机)都是没有公网IP的(只有NAT的IPV4网络)，所以需要通过跳板机进行连接。 由于Pycharm的远程连接是通过SSH进行连接的，使用SFTP协议，所以仅使用22端口即可。 ","tags":["linux","ubuntu","lxd","ssh","sftp","机房"],"title":"给机房的LXD容器配置跳板机进行连接","url":"https://www.spiritysdx.top/20240515/"},{"categories":["电脑技巧"],"date":"2024-05-14T00:00:00+00:00","lastmod":"2025-06-28T18:47:09+08:00","markdown_url":"https://www.spiritysdx.top/20240514/index.md","section":"posts","summary":"前言 相信通过前面教程，大部分问题都解决了，目前就剩一个Pycharm怎么连接远程的LXC容器运行Python项目的问题了。 (为什么不用VSCODE？因为它没有官方支持的远程开发/SSH开发的功能，只有相关的第三方插件实现了类似的功能但并不好用，所以我选择了Pycharm) ","tags":["linux","ubuntu","lxd","ssh","sftp","pycharm","机房"],"title":"在Pycharm上连接远程虚拟环境进行使用","url":"https://www.spiritysdx.top/20240514/"},{"categories":["电脑技巧"],"date":"2024-05-13T00:00:00+00:00","lastmod":"2025-06-28T18:47:09+08:00","markdown_url":"https://www.spiritysdx.top/20240513/index.md","section":"posts","summary":"前言 为什么使用LXD？ 因为宿主机不支持嵌套虚拟化，没有硬件加速没法搞PVE的虚拟机，开容器比较好。由于宿主机是Ubuntu22，自然选择LXD而不是INCUS了，因为后者在ubuntu24以上以及debian13以上才有官方支持包，其他低版本的系统只有第三方编译的包，有问题难处理。 ","tags":["linux","ubuntu","lxd","incus","https","gpu","LTS","conda","nvidia","pytorch","cuda","机房"],"title":"给机房的Ubuntu22.04安装LXD共享GPU资源","url":"https://www.spiritysdx.top/20240513/"},{"categories":["电脑技巧"],"date":"2024-05-12T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20240512/index.md","section":"posts","summary":"前言 因为机房的机器没有公网IP，所以需要跳板机进行IP映射内穿，所以需要2样东西： 1 服务器A - 服务端 - 一个有公网IPV4地址的云服务器，IPV4地址本教程假设为 x.x.x.x 1 服务器B - 客户端 - 机房内的Ubuntu22.04系统的机器 这里选用goproxy进行IP内穿 ","tags":["linux","ubuntu","goproxy","proxy","port","内穿","映射端口","机房"],"title":"给机房的Ubuntu22.04的Linux进行内穿映射端口","url":"https://www.spiritysdx.top/20240512/"},{"categories":["电脑技巧"],"date":"2024-05-11T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20240511/index.md","section":"posts","summary":"废话少说 1 sed -i \u0026#39;s/.*precedence ::ffff:0:0\\/96.*/precedence ::ffff:0:0\\/96 100/g\u0026#39; /etc/gai.conf 1 2 3 4 apt update apt install openssh-server -y systemctl enable --now ssh systemctl status ssh 1 2 3 4 5 6 7 8 9 sed -i \u0026#39;s/^#\\?Port.*/Port 22/g\u0026#39; /etc/ssh/sshd_config sed -i …","tags":["linux","ubuntu","ssh","server","机房"],"title":"给机房的Ubuntu22.04的Linux安装SSH的server","url":"https://www.spiritysdx.top/20240511/"},{"categories":["电脑技巧"],"date":"2024-05-10T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20240510/index.md","section":"posts","summary":"前置条件 1 2 3 cd /root apt update apt install curl sudo -y 安装 执行下述命令 1 export RUNNER_ALLOW_RUNASROOT=1 执行后正常安装至于./config那一步，然后不要使用官方的启动方式，而是使用守护进程启动，做到runner开机自启以及日志记录 ","tags":["linux","ubuntu","debian","github","runner","docker","root","机房"],"title":"在root权限下使用守护进程自托管Github的Runners","url":"https://www.spiritysdx.top/20240510/"},{"categories":["电脑技巧"],"date":"2024-05-09T00:00:00+00:00","lastmod":"2025-06-28T18:47:09+08:00","markdown_url":"https://www.spiritysdx.top/20240509/index.md","section":"posts","summary":"前言 前置条件以及一些碎碎念： 至少两块硬盘，一块硬盘已安装Windows11，另一块硬盘用于安装Ubuntu22.04，两块硬盘保证系统不会互相影响出奇怪的问题 ","tags":["linux","windows11","rufus","双系统","BIOS","GOST","UEFI","GPU","easyuefi","ubuntu","机房"],"title":"给机房的windows11系统安装Ubuntu22.04的Linux双系统","url":"https://www.spiritysdx.top/20240509/"},{"categories":["Golang","Python"],"date":"2024-03-28T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20240328/index.md","section":"posts","summary":"前言 之前有使用 FastAPI 开发过一套后端，但占用属实有点大，最近使用 Gin-Vue-Admin 进行重构，在构建爬虫模块的时候，发现了一些有趣的事情，Python的重定向比Golang的重定向次数要少 ","tags":["requests","gorequest","重定向"],"title":"Golang的重定向比Python的重定向次数要多","url":"https://www.spiritysdx.top/20240328/"},{"categories":["电脑技巧"],"date":"2024-01-11T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20240111/index.md","section":"posts","summary":"前言 最近在玩IPV6，遇到不少没人记录和讨论过的问题，做做记录和吐槽。 pve proxmox-ve虽然是开源的受大多数人使用的平台，但在网络管理方面还是一塌糊涂 ","tags":["linux","pve","vps","kvm","virtual","qemu","ipv6","debian","slaac","DHCPV6"],"title":"一些关于IPV6分配和使用的闲聊","url":"https://www.spiritysdx.top/20240111/"},{"categories":["机器学习","Python"],"date":"2023-11-09T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20231101/index.md","section":"posts","summary":"GNN ","tags":["机器学习","深度学习","图神经网络","GNN","GIN","CS244W"],"title":"神经网络(GNN)的表达能力","url":"https://www.spiritysdx.top/20231101/"},{"categories":["电脑技巧"],"date":"2023-10-25T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20231025/index.md","section":"posts","summary":"前言 众所周知的原因，服务器大厂的监控一直是介于保护隐私和侵犯隐私的边缘的，所以为了去除掉这些烦人的监控，我开发了下面的脚本进行一键卸载 脚本 脚本仓库：https://github.com/spiritLHLS/one-click-installation-script ","tags":["linux"],"title":"阿里云、腾讯云、甲骨文云、华为云、UCLOUD的监控卸载","url":"https://www.spiritysdx.top/20231025/"},{"categories":["机器学习","Python"],"date":"2023-10-24T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20231024/index.md","section":"posts","summary":"GAT 原文： 流程 上面这个是计算每个邻居节点和当前节点之间的重要性，下面这里是softmax这些值，计算每个节点和当前节点之间的权重，权重和为1. ","tags":["机器学习","深度学习","图神经网络","GAT","MPNN","CS244W"],"title":"GAT 详解","url":"https://www.spiritysdx.top/20231024/"},{"categories":["机器学习","Python"],"date":"2023-10-23T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20231023/index.md","section":"posts","summary":"前言 空域提出 –\u0026gt; 频域提出 –\u0026gt; 图神经网络提出 –\u0026gt; 主要回到空域发展 图表示学习 和 图神经网络 平行的两个方向，互补的方向。 MPNN 后面两个图神经网络都是基于 MPNN 进行改进的。 ","tags":["机器学习","深度学习","图神经网络","GraphSAGE","MPNN","CS244W"],"title":"GraphSAGE 详解","url":"https://www.spiritysdx.top/20231023/"},{"categories":["机器学习","Python"],"date":"2023-10-22T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20231022/index.md","section":"posts","summary":"前言 GCN 可以说是一个函数拟合器。 每个计算图就是一个样本 感受野是指神经网络中某一层输出的单元对输入的局部区域的敏感程度。实际上层数不宜过多，否则后续层数感受野会越来越相似。 ","tags":["机器学习","深度学习","图神经网络","CNN","GCN","CS244W"],"title":"GCN 详解","url":"https://www.spiritysdx.top/20231022/"},{"categories":["机器学习","Python"],"date":"2023-10-21T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20231021/index.md","section":"posts","summary":" ","tags":["机器学习","深度学习","图神经网络","GNN","CS244W"],"title":"图神经网络的前情提要","url":"https://www.spiritysdx.top/20231021/"},{"categories":["机器学习","Python"],"date":"2023-10-20T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20231020/index.md","section":"posts","summary":" ","tags":["机器学习","深度学习","图神经网络","CS244W"],"title":"深度学习前情提要","url":"https://www.spiritysdx.top/20231020/"},{"categories":["电脑技巧"],"date":"2023-09-29T00:00:00+00:00","lastmod":"2023-11-10T17:19:58+08:00","markdown_url":"https://www.spiritysdx.top/20211231/index.md","section":"posts","summary":"给轻量应用腾讯云服务器加上IPV6地址隧道 这里使用6in4方法解决宿主机本身没有IPV6地址的问题。 给宿主机附加免费的IPV6地址段 有的机器本身没有IPV6的/64子网，这里给出一个方法免费附加IPV6的子网。 这里是使用6in4方法解决宿主机本身没有IPV6地址的问题。 ","tags":["linux","ipv6"],"title":"给轻量应用腾讯云服务器或阿里云服务器加上IPV6地址隧道(2023)","url":"https://www.spiritysdx.top/20211231/"},{"categories":["机器学习","Python"],"date":"2023-09-28T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230928/index.md","section":"posts","summary":"前言 ","tags":["机器学习","深度学习","图神经网络","CS244W"],"title":"半监督节点分类","url":"https://www.spiritysdx.top/20230928/"},{"categories":["机器学习","Python"],"date":"2023-09-27T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230927/index.md","section":"posts","summary":"PageRank节点重要度 在NetworkX中，计算有向图节点的PageRank节点重要度。 参考资料 networkx官方教程：https://networkx.org/documentation/stable/tutorial.html nx.Graph https://networkx.org/documentation/stable/reference/classes/graph.html#networkx.Graph 给图、节点、连接添加属性 …","tags":["机器学习","深度学习","图神经网络","PageRank","CS244W"],"title":"PageRank实战-西游记人物节点重要程度","url":"https://www.spiritysdx.top/20230927/"},{"categories":["机器学习","Python"],"date":"2023-09-26T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230926/index.md","section":"posts","summary":"Links as votes 使用In-coming links作为投票，也就是别人引用了我，即一个节点被多少其他节点指向，那么这个节点被多少用户点击的概率就越大。 pagerank假设 In-Links 之间也是不一样，重要网站引用我和无名小将引用我肯定前者更重要。 ","tags":["机器学习","深度学习","图神经网络","PageRank","CS244W"],"title":"PageRank","url":"https://www.spiritysdx.top/20230926/"},{"categories":["机器学习","Python"],"date":"2023-09-25T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230925/index.md","section":"posts","summary":"基于原文作的个人笔记，在此展出 建议按住ctrl然后滑动鼠标滑轮放大查看，笔记做的很小，图片分辨率很高，放大也不会糊 原文：Combinatorial Optimization with Graph Convolutional Networks and Guided Tree Search 原文链接：https://arxiv.org/abs/1810.10659 ","tags":["机器学习","深度学习","图神经网络","树搜索","组合优化","局部搜索","图缩减"],"title":"用图卷积和树搜索解决组合优化问题(含笔记)","url":"https://www.spiritysdx.top/20230925/"},{"categories":["机器学习","Python"],"date":"2023-09-24T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230924/index.md","section":"posts","summary":"前言 DeepWalk 仅能反映相邻节点的社群相似信息，无法反映节点的功能角色相似的信息，因为DeespWalk仅使用连接信息(结构信息)学习。 有偏随机游走 ","tags":["机器学习","深度学习","图神经网络","Node2Vec","CS244W"],"title":"Node2Vec","url":"https://www.spiritysdx.top/20230924/"},{"categories":["机器学习","Python"],"date":"2023-09-23T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230923/index.md","section":"posts","summary":"参考资料 https://www.analyticsvidhya.com/blog/2019/11/graph-feature-extraction-deepwalk/ https://github.com/prateekjoshi565/DeepWalk 安装工具包 1 !pip install networkx gensim pandas numpy tqdm scikit-learn matplotlib Collecting networkx Downloading networkx-3.1-py3-none-any.whl (2.1 MB) …","tags":["机器学习","深度学习","图神经网络","DeepWalk","CS244W"],"title":"DeepWalk实战-维基百科词条图嵌入可视化","url":"https://www.spiritysdx.top/20230923/"},{"categories":["机器学习","Python"],"date":"2023-09-22T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230922/index.md","section":"posts","summary":"前言 深度学习思想用于图嵌入的开山之作 - DeepWalk 学习路径： Random Walk 是一种遍历图的基本方法，而 Deep Walk 则是一种利用 Random Walk 生成的序列来学习节点嵌入的具体方法。 ","tags":["机器学习","深度学习","图神经网络","DeepWalk","CS244W"],"title":"DeepWalk","url":"https://www.spiritysdx.top/20230922/"},{"categories":["机器学习","Python"],"date":"2023-09-21T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230921/index.md","section":"posts","summary":"如何嵌入 全图进行编码嵌入 引入虚拟节点 匿名随机游走 https://arxiv.org/pdf/1805.11921.pdf 只要是新节点就标记新的一号，认号不认原来的节点标识。 ","tags":["机器学习","深度学习","图神经网络","DeepWalk","Node2Vec","CS244W"],"title":"嵌入整张图","url":"https://www.spiritysdx.top/20230921/"},{"categories":["机器学习","Python"],"date":"2023-09-20T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230920/index.md","section":"posts","summary":"前言 矩阵分解和随机游走在数学上意义是一样的。 矩阵分解计算 矩阵分解这里可能解析解不能直接求得，且可能不唯一，所以实际计算用的数值解。 ","tags":["机器学习","深度学习","图神经网络","DeepWalk","Node2Vec","CS244W"],"title":"矩阵分解角度 - 图嵌入(节点嵌入)和随机游走","url":"https://www.spiritysdx.top/20230920/"},{"categories":["机器学习","Python"],"date":"2023-09-19T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230919/index.md","section":"posts","summary":"前言 本文讲述映射成D维向量的基于随机游走方法。 学习路径： 图嵌入 - Node Embedding 把节点映射为一个D维向量。 ","tags":["机器学习","深度学习","图神经网络","MiniBatch","RandomWalk","Node2Vec","CS244W"],"title":"Random Walk \u0026 Node2vec - 图表示学习起始","url":"https://www.spiritysdx.top/20230919/"},{"categories":["机器学习","Python"],"date":"2023-09-18T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230918/index.md","section":"posts","summary":"北京上海地铁站图数据挖掘 上海、北京地铁站点图数据挖掘，计算地铁站点的最短路径、节点重要度、集群系数、连通性。 导入工具包 1 2 3 4 5 import networkx as nx import pandas as pd import matplotlib.pyplot as plt import matplotlib.colors as mcolors %matplotlib inline 可视化辅助函数 1 2 3 4 5 6 7 8 9 10 11 12 13 14 def draw(G, pos, measures, …","tags":["机器学习","深度学习","特征工程","图神经网络","networkx","CS244W"],"title":"NetworkX实战","url":"https://www.spiritysdx.top/20230918/"},{"categories":["机器学习","Python"],"date":"2023-09-17T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230917/index.md","section":"posts","summary":"PageRank节点重要度 在NetworkX中，计算有向图节点的PageRank节点重要度。 参考资料 networkx官方教程：https://networkx.org/documentation/stable/tutorial.html nx.Graph https://networkx.org/documentation/stable/reference/classes/graph.html#networkx.Graph 给图、节点、连接添加属性 …","tags":["机器学习","深度学习","特征工程","图神经网络","networkx","CS244W"],"title":"NetworkX实用模板案例","url":"https://www.spiritysdx.top/20230917/"},{"categories":["机器学习","Python"],"date":"2023-09-16T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230916/index.md","section":"posts","summary":"nx.draw可视化函数 创建4x4网格图 1 G = nx.grid_2d_graph(4, 4) 原生可视化 1 pos = nx.spring_layout(G, seed=123) 1 nx.draw(G, pos) 不显示节点 1 nx.draw(G, pos, node_size=0, with_labels=False) 设置颜色 1 len(G.edges()) 24 1 2 3 4 5 6 7 8 9 10 11 nx.draw( G, pos, node_color=\u0026#39;#A0CBE2\u0026#39;, # 节点颜色 …","tags":["机器学习","深度学习","特征工程","图神经网络","networkx","CS244W"],"title":"NetworkX可视化","url":"https://www.spiritysdx.top/20230916/"},{"categories":["机器学习","Python"],"date":"2023-09-15T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230915/index.md","section":"posts","summary":"前言 当然，还是习惯 linux + jupyterlab 的组合，当然还是基于 https://www.spiritlhl.net/case/case1.html#%E4%B8%80%E9%94%AE%E5%AE%89%E8%A3%85jupyter%E7%8E%AF%E5%A2%83 构建基础环境 安装 由于是非国内服务器，无需清华镜像 1 2 !pip3 install numpy pandas matplotlib tqdm networkx # -i https://pypi.tuna.tsinghua.edu.cn/simple 验证是否安装 …","tags":["机器学习","深度学习","特征工程","图神经网络","networkx","CS244W"],"title":"NetworkX基本操作","url":"https://www.spiritysdx.top/20230915/"},{"categories":["机器学习","Python"],"date":"2023-09-14T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230914/index.md","section":"posts","summary":"前言 需要提取特征反映全图结构特定 依然是在数数字 Bag-of-Words - (BoW) 数节点是否存在，存在为0，不存在为1： 数节点度数为xx的节点有几个： ","tags":["机器学习","深度学习","特征工程","图神经网络","核方法","Bow","gklearn","CS244W"],"title":"传统图机器学习和图特征工程(全图层面的特征工程)","url":"https://www.spiritysdx.top/20230914/"},{"categories":["机器学习"],"date":"2023-09-13T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230913/index.md","section":"posts","summary":"前言 通过已知连接补全未知连接 连接预测 连接的特征 两节点的距离特征 两节点的最短路径长度： ","tags":["机器学习","深度学习","特征工程","图神经网络","CS244W"],"title":"传统图机器学习和图特征工程(连接层面的特征工程)","url":"https://www.spiritysdx.top/20230913/"},{"categories":["机器学习"],"date":"2023-09-12T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230912/index.md","section":"posts","summary":"前言 这里就是解决之前生成D维向量的问题，使用人工设置特征，后续会使用GNN也就是图神经网络自动学习特征而不再需要人工设置特征了，这块等同于是被替代掉的工作，但由于后续的自动特征提取还是起源于这里的，所以还是需要了解一下。 ","tags":["机器学习","深度学习","特征工程","图神经网络","CS244W"],"title":"传统图机器学习和图特征工程(节点层面的特征工程)","url":"https://www.spiritysdx.top/20230912/"},{"categories":["机器学习"],"date":"2023-09-11T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230911/index.md","section":"posts","summary":"图基础 本体图是导入图之前就应该设计好的，本体图和具体图的关系类同类和实例的关系。如何设计本体图取决于将来的图的用途，比如图的用途是分析，那么本体图的设计应该包含所有图的属性，比如节点和边的属性，以及节点和边的类型，比如节点是用户还是物品，边是用户对物品感兴趣还是用户对用户感兴趣。 ","tags":["机器学习","深度学习","图神经网络","CS244W"],"title":"图的基本表示","url":"https://www.spiritysdx.top/20230911/"},{"categories":["机器学习","Python"],"date":"2023-09-10T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230910/index.md","section":"posts","summary":"传统机器学习 默认二者独立同分布，只需要拟合决策边界分类或拟合回归的曲线即可。 现代神经网络 斯坦福CS的相关课程： 网络类型 数据类型 课程 全连接神经网络 表格 无 卷积神经网络 图像 CS231N 循环神经网络、Transformer 文本语音带序列 CS224N 图神经网络 图数据 CS244W 复杂的图结构 任意尺寸输入 没有固定的节点顺序和参考锚点 动态变化，多模态特征 表示学习 - 图嵌入 - node embedding 把一个复杂的图节点表示为一个d维向量，能充分表示原始数据的语义。 ","tags":["机器学习","深度学习","图神经网络","Python","CS244W"],"title":"图神经网络基础","url":"https://www.spiritysdx.top/20230910/"},{"categories":["机器学习","Python"],"date":"2023-09-09T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230909/index.md","section":"posts","summary":"前言 所有资源均免费，遇到付费推荐勿要付费 方向是 图神经网络 和 强化学习 的学习路径 基础知识 白月黑羽的网站： https://www.byhy.net/ 适合学习Python基础知识，有对应的B站讲解视频 他的自动化方面课程讲的也很好，爬虫中的selenium，以及Python的自动化测试，都算是他很好的课程了。(属于题外话了，这些东西在深度学习中用不到，没时间的别浪费时间看) ","tags":["机器学习","深度学习","强化学习","图神经网络","Python","CS244W"],"title":"Python相关的深度学习的学习路径","url":"https://www.spiritysdx.top/20230909/"},{"categories":["机器学习","Python"],"date":"2023-09-08T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230908/index.md","section":"posts","summary":"前言 Actor-Critic Methods 结合了价值学习和策略学习，同时训练了两个神经网络。 Actor 网络用于产生策略，Critic 网络用于评估策略。 目标 ① 更新策略网络Π的参数，是为了增大状态价值V的值，要用价值网络q进行打分来训练。 ","tags":["机器学习","深度学习","强化学习","Python"],"title":"Actor-Critic Methods","url":"https://www.spiritysdx.top/20230908/"},{"categories":["机器学习","Python"],"date":"2023-09-07T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230907/index.md","section":"posts","summary":"用策略函数指导动作 使用策略函数随机抽样得到动作。 近似策略函数 由于实际的策略函数无法得到，需要用各种方式去近似策略函数，所以这里可以使用神经网络去近似实际的策略函数，记作policy network。 ","tags":["机器学习","深度学习","强化学习","Python"],"title":"Policy-Based learning","url":"https://www.spiritysdx.top/20230907/"},{"categories":["机器学习","Python"],"date":"2023-09-06T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230906/index.md","section":"posts","summary":"寻找最佳的Q值函数 实际并不知道最佳的Q值函数，需要使用神经网络 Q(s,a;w) 来近似最佳的Q值函数。 实际流程大致为当前状态转换为矩阵后，通过卷积层提取特征向量，再通过全连接层得到Q值向量，此时的Q值向量每一个元素代表某一个动作的得分。 ","tags":["机器学习","深度学习","强化学习","Python","DQN"],"title":"Deep Q-Network (DQN) (Value-Based learning)","url":"https://www.spiritysdx.top/20230906/"},{"categories":["机器学习","Python"],"date":"2023-09-05T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230905/index.md","section":"posts","summary":"State 状态，即状态空间，表示环境中的当前状态。 Action \u0026amp;\u0026amp; Agent 动作，即动作空间，表示在当前状态下，执行的动作。 动作由谁做的就是Agent，即智能体。 Policy Π 策略，即策略空间，表示在当前状态下，智能体可以采取的动作。 ","tags":["机器学习","深度学习","强化学习","Python","gym"],"title":"强化学习术语翻译","url":"https://www.spiritysdx.top/20230905/"},{"categories":["电脑技巧"],"date":"2023-08-29T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230829/index.md","section":"posts","summary":"前言 观望了全网的Docker启用IPV6的方法，要么是Docker版本更替法子不通了，要么是没说明一些前置条件的细节，导致方法也用不了，所以这里记录一下我走通的方法，一个兼容高低版本Docker和不同网络环境的方法 ","tags":["linux","docker","nat","radvd","ndppd","ndpresponder","ifupdown","shell","debian","ipv6"],"title":"为Docker配置启用IPV6网络并配置给容器自动分配IPV6地址(2023最新)","url":"https://www.spiritysdx.top/20230829/"},{"categories":["机器学习","Python"],"date":"2023-08-23T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230823/index.md","section":"posts","summary":"前言 老板下指示复现两篇文章，这是其中一篇 https://arxiv.org/pdf/2205.14105v1.pdf 文章的原理什么的已经大部分明白了但仍然有部分懂，故而做下记录，以备后续复现或深入了解 原始数据 ER40/BA40到ER500/BA500 https://ojs.aaai.org/index.php/AAAI/article/download/5723/5579 https://journals.aps.org/rmp/abstract/10.1103/RevModPhys.74.47 分别命名为ER和BA数据集 ","tags":["机器学习","深度学习","jupyter","pytorch","conda","Python","GCN","RNN"],"title":"高效探索学习解决组合图分区问题(基于强化学习的优化算法)","url":"https://www.spiritysdx.top/20230823/"},{"categories":["机器学习","Python"],"date":"2023-08-16T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230816/index.md","section":"posts","summary":"前言 两篇文章的主体解析没有涉及作者进行模型比较的部分，这里主要解决一下该部分 由于两篇文章都涉及该方法的比较，所以重头戏是DQN以及其衍生的一些变体，还有部分别的模型，是需要提前了解的。 ","tags":["机器学习","深度学习","jupyter","DQN","RL-GNN","SOTA"],"title":"图优化问题经常拿来比较的一些模型","url":"https://www.spiritysdx.top/20230816/"},{"categories":["机器学习","Python"],"date":"2023-08-12T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230812/index.md","section":"posts","summary":"原地操作 使用 1 2 3 4 a = 1 b = 1 a += b print(a) # 结果是2 而不是使用 1 a = a + b 好处是内存不会复制扩展，只使用a和b的内存运算 1 2 3 4 import numpy as np X = np.arange(12).reshape(3, 4) Y = np.array([[2, 1, 4, 3], [1, 2, 3, 4], [4, 3, 2, 1]]) np.concatenate([X, Y], axis=0), np.concatenate([X, Y], axis=1) 检测内存是否一致 …","tags":["机器学习","深度学习","jupyter","Python"],"title":"Python加速科学运算的一些小技巧","url":"https://www.spiritysdx.top/20230812/"},{"categories":["机器学习","Python"],"date":"2023-08-08T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230808/index.md","section":"posts","summary":"前言 老板下指示复现两篇文章，这是其中一篇 https://arxiv.org/pdf/1810.10659.pdf 文章的原理什么的已经大致明白了但仍然有小部分不懂，故而做下记录，以备后续复现或深入了解 原始数据 Training Data https://www.cs.ubc.ca/~hoos/SATLIB/benchm.html Testing Data SAT Competition 2017 …","tags":["机器学习","深度学习","jupyter","dgl","pytorch","conda","Python","GCN","Gurobi","CNF","GR","LS"],"title":"通过GCN生成概率图引导树搜索解决图的组合优化问题","url":"https://www.spiritysdx.top/20230808/"},{"categories":["机器学习","Python"],"date":"2023-07-27T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230727/index.md","section":"posts","summary":"由于之前有写过一键安装jupyter的shell脚本，所以这里只需要找一个服务器就够了 https://github.com/spiritLHLS/one-click-installation-script#%E4%B8%80%E9%94%AE%E5%AE%89%E8%A3%85jupyter%E7%8E%AF%E5%A2%83 1 curl -L …","tags":["机器学习","深度学习","jupyter","jupyterlab","d2l","conda","Python"],"title":"深度学习环境安装(李沐老师相关)","url":"https://www.spiritysdx.top/20230727/"},{"categories":["机器学习"],"date":"2023-07-18T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230714/index.md","section":"posts","summary":"一、命题逻辑与真值函数 命题逻辑：（又称命题演算、布尔逻辑）是最简单的一种形式逻辑系统。 主要研究对象：命题（常用p,q,r…代表任意命题即命题变元）每个命题可能为真，也可能为假（通常用1/0或T/F或T/⊥表示）。 ","tags":["机器学习","深度学习","组合优化"],"title":"命题逻辑","url":"https://www.spiritysdx.top/20230714/"},{"categories":["电脑技巧"],"date":"2023-06-11T00:00:00+00:00","lastmod":"2023-10-22T12:44:14+08:00","markdown_url":"https://www.spiritysdx.top/20230611/index.md","section":"posts","summary":"要通过WiFi共享D盘给其他电脑，可以使用以下方法： 1.创建共享文件夹：首先，需要在D盘上创建一个共享文件夹。右键单击D盘上的文件夹，选择\u0026#34;属性\u0026#34;，然后切换到\u0026#34;共享\u0026#34;选项卡。点击\u0026#34;高级共享\u0026#34;，勾选\u0026#34;共享此文件夹\u0026#34;选项，并为文件夹指定一个共享名称。(或者直接就右键D盘，打开属性) ","tags":["windows"],"title":"共享D盘给同一局域网下的其他电脑(WiFi)","url":"https://www.spiritysdx.top/20230611/"},{"categories":["电脑技巧"],"date":"2023-05-06T00:00:00+00:00","lastmod":"2023-10-22T12:44:14+08:00","markdown_url":"https://www.spiritysdx.top/20230506/index.md","section":"posts","summary":"PVE 感谢 Proxmox VE 的免费订阅支持 原始仓库：https://github.com/spiritLHLS/pve 说明文档 国内(China)： virt.spiritlhl.net 国际(Global)： www.spiritlhl.net 说明文档中 Proxmox VE 分区内容 https://github.com/oneclickvirt/kvm_images 为对应虚拟机镜像仓库 前言 国内服务器请使用国内命令，国际服务器请使用国际命令 ","tags":["linux","pve","vps","kvm","virtual","qemu","nat","iptables","shell","debian"],"title":"一键安装PVE并一键开设KVM虚拟化的NAT服务器-带内外网端口转发","url":"https://www.spiritysdx.top/20230506/"},{"categories":["python"],"date":"2023-04-22T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20230422/index.md","section":"posts","summary":"近期码代码的个人心得 尊重他人命运，放弃助人情节，帮得了一次往往意味着帮无数次 不要让不懂行的人改你的代码，改了最好别再接手代码，往往写的越来越屎，接手了意味着有锅的话你得背，很难辨清 ","tags":["python"],"title":"近期码代码的个人心得","url":"https://www.spiritysdx.top/20230422/"},{"categories":["电脑技巧"],"date":"2023-02-22T00:00:00+00:00","lastmod":"2023-10-22T12:44:14+08:00","markdown_url":"https://www.spiritysdx.top/20230222/index.md","section":"posts","summary":"前言 该方法的前提： 商家使用PVE虚拟化WIN服务器 WIN服务器可配置使用Network进行启动 启动时可使用NO VNC操控IPXE命令 已知符合条件的服务器：NETFRONT ","tags":["linux","pve","vps","kvm","virtual","qemu","ipxe","hongkong","dd"],"title":"将WIN系统DD为Linux系统","url":"https://www.spiritysdx.top/20230222/"},{"categories":["python","数学建模","机器学习"],"date":"2022-12-30T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20221230/index.md","section":"posts","summary":"只给出个人完成的代码部分与一些简单的结果和说明 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 from sklearn.decomposition import PCA from sklearn.ensemble import RandomForestClassifier as RFC from sklearn.neighbors import KNeighborsClassifier as KNN from sklearn.model_selection import cross_val_score from sklearn …","tags":["python","PCA","LLE","T-SNE","机器学习","特征工程"],"title":"关于手写数字识别的特征工程","url":"https://www.spiritysdx.top/20221230/"},{"categories":["python","爬虫"],"date":"2022-12-10T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20221212/index.md","section":"posts","summary":"解决上一篇文章的一些问题 解决方法： https://github.com/wkeeling/selenium-wire/issues/622 下载证书 https://github.com/wkeeling/selenium-wire/raw/master/seleniumwire/ca.crt 将该 ca.crt 添加到受信任的证书中 google浏览器打开chrome://settings/security 点击管理证书，选择受信任的的根证书颁发机构分区，选择导入 ","tags":["python","selenium","chromedriver","geckodriver","edgedriver","socks5","seleniumwire","selenium-wire","代理"],"title":"在seleniumwire中解决Not Secure问题","url":"https://www.spiritysdx.top/20221212/"},{"categories":["python","爬虫"],"date":"2022-12-09T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200516/index.md","section":"posts","summary":"前言 需要下载Chrome或Firefox的driver，Chrome内核81.440与Firefox内核74.0下载链接如下： Firefox Chrome 其他版本请在搜索引擎查找，本篇使用该版本，注意，driver下载后需要配置对应内核的游览器，电脑本身需要有该内核的游览器。 ","tags":["python","爬虫"],"title":"Selenium的web自动化操作02(基本语法)","url":"https://www.spiritysdx.top/20200516/"},{"categories":["python","爬虫"],"date":"2022-12-09T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20221211/index.md","section":"posts","summary":"前言 爬虫过程中出现了cloudflare的高风险质询验证阻拦或者recaptcha的阻拦，想要破除验证要么使用干净的代理保证低风险爬虫，要么使用hcaptcha求解器求解，后者需要部署的东西过于复杂，为了提高效率，本文介绍第一种方法破除质询验证 ","tags":["python","selenium","hcaptcha","recaptcha","captcha","验证码","chromedriver","geckodriver","edgedriver","socks5","中间件","seleniumwire","selenium-wire","python-proxy","pproxy","goproxy","代理"],"title":"解决selenium爬虫加密代理问题(含socks5等一切代理)","url":"https://www.spiritysdx.top/20221211/"},{"categories":["电脑技巧"],"date":"2022-12-06T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20221206/index.md","section":"posts","summary":"为openvz或kvm架构的linux服务器增加swap分区 addswap 更新时间：2022.12.05 为openvz或kvm架构的linux服务器增加swap分区，请确保在root权限下使用 1 2 sudo -i curl -L https://raw.githubusercontent.com/spiritLHLS/addswap/main/addswap.sh -o addswap.sh \u0026amp;\u0026amp; chmod +x addswap.sh \u0026amp;\u0026amp; bash addswap.sh 已增加openvz架构重启swap自动添加。 ","tags":["linux","shell","server","vps","kvm","openvz","virtual","swap"],"title":"为openvz或kvm架构的linux服务器增加swap分区","url":"https://www.spiritysdx.top/20221206/"},{"categories":["电脑技巧"],"date":"2022-12-05T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20221205/index.md","section":"posts","summary":"使用LXD对服务器进行LXC容器切分(下面简称母鸡开小鸡) Linux母鸡开小鸡，一键多开NAT小鸡，一键LXC虚拟化，一键多开服务器，多开容器，一键多开NAT小鸡，一键多开NAT服务器 上述需求都得到了解决 ","tags":["linux","lxd","lxc","shell","server","nat","vps","kvm","virtual","vnstat"],"title":"使用LXD对服务器进行LXC容器切分","url":"https://www.spiritysdx.top/20221205/"},{"categories":["python","爬虫"],"date":"2022-11-03T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20221103/index.md","section":"posts","summary":"前言 前面有破解谷歌验证码，现在来破解相对简单的图片验证码。(数字以及中英混合) 插件过验证码 需要加载谷歌插件AutoVerify 这种识别是为了过简单的文字+数字的验证图片用的方法。 ","tags":["python","selenium","recaptcha","captcha","验证码"],"title":"免费破解图片验证码(数字或中英混合)(附代码)(2022)","url":"https://www.spiritysdx.top/20221103/"},{"categories":["python","爬虫"],"date":"2022-11-02T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20221102/index.md","section":"posts","summary":"前言 破解谷歌验证码，其实并不需要用高深的法子，只需要借助一些免费资源，免费插件，就能达到很好的效果。 网上现在流传的几乎都说Google验证码ReCaptchav3得用深度学习，或者第三方收费打码网站，又或者抓住某些验证漏洞才能过，实际上，并不需要。这里我先从普通的验证码开始，介绍我破解Google验证码的思路以及个人成功案例。(全网独家，目前我各大搜索引擎找遍了都没有的，如需转发请注明本文来源) ","tags":["python","selenium","shadow-root","reCAPTCHA-V3","验证码"],"title":"Python破解Google验证码ReCaptchav3的成功案例(附代码)(免费)(2022)","url":"https://www.spiritysdx.top/20221102/"},{"categories":["python","数学建模"],"date":"2022-02-24T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20220224/index.md","section":"posts","summary":"前言 美赛论文对应主体代码部分，吐槽一句美赛居然不收代码也是离谱。 论文配套的代码均为本人编写，运行无问题。 感谢组员组长的配合论文撰写。 正文 1 2 3 4 5 6 7 8 9 import numpy as np import pandas as pd import re,math import matplotlib.pyplot as plt from scipy.optimize import linprog np.set_printoptions(suppress=True) data1 = …","tags":["python"],"title":"(MCM/ICM)比特币和黄金组合投资策略的主体代码部分","url":"https://www.spiritysdx.top/20220224/"},{"categories":["python","数学建模"],"date":"2022-02-23T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20220226/index.md","section":"posts","summary":"感谢组员组长的配合论文撰写。 1 Introduction Background and restatement of the problem 在如今金融迅猛发展的大背景下，交易员可以对众多的金融投资产品进行选择，并对其投资的波动性资产进行交易，以收获期望的回报。其中黄金和比特币为本文所选取的金融资产。比特币每天都进行交易而黄金只在开市日进行交易，且在交易中交易员需要支付对应比例的佣金。 ","tags":["python"],"title":"(MCM/ICM)比特币和黄金组合投资策略","url":"https://www.spiritysdx.top/20220226/"},{"categories":["python","数学建模"],"date":"2022-02-23T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20220225/index.md","section":"posts","summary":"前言 美赛论文对应策略代码部分，吐槽一句美赛居然不收代码也是离谱。 论文配套的代码均为本人编写，运行无碍。 感谢组员组长的配合论文撰写。 正文 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 import numpy as np import pandas as pd import re,math import matplotlib.pyplot as plt np.set_printoptions(suppress=True) B = pd.read_csv(r\u0026#39;B.csv\u0026#39;) # B H = …","tags":["python"],"title":"(MCM/ICM)比特币和黄金组合投资策略的策略代码部分","url":"https://www.spiritysdx.top/20220225/"},{"categories":["python","数学建模"],"date":"2022-02-22T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20220222/index.md","section":"posts","summary":"前言 美赛论文，由于不能在md中方便的插入图片与数学公式，所以我选择删掉对应的地方。 英文译制版本。 论文配套的代码均为本人编写，感谢组员组长的配合论文撰写。 ","tags":["python"],"title":"(MCM/ICM)The Best Investment Strategies For Gold And Bitcoin","url":"https://www.spiritysdx.top/20220222/"},{"categories":["python","数学建模"],"date":"2022-02-21T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20220221/index.md","section":"posts","summary":"背景： 市场交易者频繁买卖波动性资产，目标是最大化其总回报。每次买卖通常都会有佣金。两种这样的资产是黄金和比特币。 图 1：黄金每日价格，每金衡盎司美元。资料来源：伦敦金银市场协​​会，2021 年 9 月 11 日 ","tags":["python"],"title":"(MCM/ICM)2022C题比特币与黄金投资问题汉译题目","url":"https://www.spiritysdx.top/20220221/"},{"categories":["python","数学建模"],"date":"2022-02-13T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20220213/index.md","section":"posts","summary":"前言 很水的比赛，官方求着收论文，说是保底有钱。。。。 英文译制版本。 感谢组员组长的配合论文撰写。 正文 Suzhou Carbon Neutral Circular Development Creative Plan Problem Analysis 1.1 Background brief 1.1.1 Problem introduction In recent years, Suzhou has faced various new problems and challenges, such as economic transformation, …","tags":["python"],"title":"Suzhou Carbon Neutral-Circular Development Creative Plan","url":"https://www.spiritysdx.top/20220213/"},{"categories":["python","数学建模"],"date":"2022-02-12T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20220212/index.md","section":"posts","summary":"前言 很水的比赛，官方求着收论文，说是保底有钱。。。。 感谢组员组长的配合论文撰写。 正文 问题分析 背景简述 问题引入 近年来，苏州由于面临经济的转型、低碳发展与社会进步协调性降低、经济发展面临的不确定性增加等各种新问题、新挑战，给碳中和目标的实现带来的冲击与挑战。而一次能源过度依赖外地调入、资源禀赋不足、能源消费多以煤炭等传统化石燃料为主、新能源发展乏力且占市场份额较小等诸多因素的影响，直接或间接的造成了苏州整体的碳排放量较高的情况。 ","tags":["python"],"title":"苏州市碳中和-循环发展创意方案书","url":"https://www.spiritysdx.top/20220212/"},{"categories":["python","爬虫"],"date":"2022-02-06T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20220206/index.md","section":"posts","summary":"直接使用原始图片做验证码识别正确率较低，使用增强技术后能大大提高识别率 黑底填充白底 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 from PIL import Image import os def Convert(filename): \u0026#34;\u0026#34;\u0026#34; 将图像中白色像素转变为黑色像素 \u0026#34;\u0026#34;\u0026#34; img = Image.open(os.getcwd()+ \u0026#34;\\\\\u0026#34;+ filename) img = img.convert(\u0026#34;RGBA\u0026#34;) pixdata = …","tags":["python","recaptcha","captcha","验证码"],"title":"图片验证码增强技术(提高识别正确率)","url":"https://www.spiritysdx.top/20220206/"},{"categories":["电脑技巧"],"date":"2022-01-20T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20220120/index.md","section":"posts","summary":"2022.1.16 Euserv激活账号以及安装机子需要钱了，无法白嫖了。。。 但是有别的机子有类似机制能白嫖，详情请看 https://github.com/spiritLHL/Hang-up-items 原有已经嫖到的自动续期脚本还是有效的。 前言 2022.1.5更新图片链接以及部分说明 ","tags":["linux"],"title":"白嫖永久的免费VPS云服务器(2022.1.16已报废)","url":"https://www.spiritysdx.top/20220120/"},{"categories":["电脑技巧"],"date":"2022-01-19T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20220119/index.md","section":"posts","summary":"前言 因为青龙拉取部分作者仓库通过https://ghproxy.com/代理拉取频繁，导致了请求量比较大的仓库被该代理加速作者封禁。所以可以自建代理的链接替换默认的https://github.com/，自己用自己的反代不会封禁，自建代理实现ql repo或ql raw拉取实时的GitHub仓库或文件。 ","tags":["linux"],"title":"自建青龙代理拉取Github仓库或文件","url":"https://www.spiritysdx.top/20220119/"},{"categories":["电脑技巧"],"date":"2022-01-17T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20220117/index.md","section":"posts","summary":"前言 事先声明，本文为个人内网穿透经验，可能不具备复刻性质，自行查验。 前期配置 A 腾讯云的VPS一台(配置带宽决定你的上限) — 我的是Ubuntu，内核amd64 ","tags":["linux"],"title":"本地ARM机子+云服务vps指定端口内穿+开机自启=外网访问本地机子","url":"https://www.spiritysdx.top/20220117/"},{"categories":["python","数学建模"],"date":"2021-10-03T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20211003/index.md","section":"posts","summary":"前言 论文配套的代码均为本人编写，第三四问存在部分代码问题，需要改动细节部分才能运行成功，后面两问的代码时间成本很高，长则数小时，短则十几二十分钟，自己跑酌情修改迭代次数，转换赛题要求的填表格式得通过最后部分的代码才能转换。 ","tags":["python"],"title":"2021年全国大学生数学建模竞赛C题省一等奖代码部分","url":"https://www.spiritysdx.top/20211003/"},{"categories":["python","数学建模"],"date":"2021-10-02T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20211002/index.md","section":"posts","summary":"前言 博客展示的是机器转义的markdown格式的，部分内容无法正常展示，完整内容详见博客相关资源。 论文配套的代码均为本人编写，第三四问存在部分代码问题，需要改动细节部分才能运行成功，后面两问的代码时间成本很高，长则数小时，短则十几二十分钟，自己跑酌情修改迭代次数，转换赛题要求的填表格式得通过最后部分的代码才能转换。 ","tags":["python"],"title":"2021年全国大学生数学建模竞赛C题省一等奖论文部分","url":"https://www.spiritysdx.top/20211002/"},{"categories":["python","数学建模","机器学习"],"date":"2021-07-28T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210728/index.md","section":"posts","summary":"前言 逻辑回归的一个小案例，使用sklearn自带的一个数据 代码 1 2 3 4 5 6 from sklearn.linear_model import LogisticRegression as LR #逻辑回归 from sklearn.datasets import load_breast_cancer #导入乳腺癌数据集 import numpy as np import matplotlib.pyplot as plt from sklearn.model_selection import train_test_split # 导入分测试与 …","tags":["python","神经网络","机器学习"],"title":"逻辑回归1-乳腺癌数据案例","url":"https://www.spiritysdx.top/20210728/"},{"categories":["python","数学建模","机器学习"],"date":"2021-07-27T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210727/index.md","section":"posts","summary":"前言 数据集合来自kaggle 链接：https://www.kaggle.com/c/digit-recognizer/data 其中test和train的csv数据集为所需数据集。 ","tags":["python","神经网络","机器学习"],"title":"随机森林在乳腺癌数据上的调参","url":"https://www.spiritysdx.top/20210727/"},{"categories":["电脑技巧"],"date":"2021-07-26T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210726/index.md","section":"posts","summary":"删除指定端口的所有进程 1 sudo kill -9 $(lsof -i:端口号 -t) 删除指定名称的所有进程 1 sudo kill -9 $(pidof 名字) 查看指定端口占用情况 1 sudo lsof -i:端口号 删除指定pid的进程 1 sudo kill -9 pid号 ","tags":["linux"],"title":"删除进程操作","url":"https://www.spiritysdx.top/20210726/"},{"categories":["电脑技巧"],"date":"2021-07-25T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210725/index.md","section":"posts","summary":"docker-compose搭建 wordpress 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 version: \u0026#39;3.3\u0026#39; services: db: image: mysql:latest volumes: - /data/wordpress_db:/var/lib/mysql restart: always environment: MYSQL_ROOT_PASSWORD: rootwordpress MYSQL_DATABASE: …","tags":["github","linux","docker"],"title":"Okteto常用项目","url":"https://www.spiritysdx.top/20210725/"},{"categories":["数学建模","python","机器学习"],"date":"2021-07-24T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210724/index.md","section":"posts","summary":"前言 使用的相关数据链接如下： 点我跳转 代码 连续值预处理 1 2 3 4 5 6 7 8 9 10 import pandas as pd data = pd.read_csv(\u0026#34;Narrativedata.csv\u0026#34; ,index_col=0 )#index_col=0将第0列作为索引，不写则认为第0列为特征 data.loc[:,\u0026#34;Age\u0026#34;] = data.loc[:,\u0026#34;Age\u0026#34;].fillna(data.loc[:,\u0026#34;Age\u0026#34;].median()) #将年龄二值化 data_2 = data.copy() data_2.info() \u0026lt;class …","tags":["python","数学建模","神经网络","机器学习"],"title":"预处理3-连续值与特征选取","url":"https://www.spiritysdx.top/20210724/"},{"categories":["数学建模","python","机器学习"],"date":"2021-07-23T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210723/index.md","section":"posts","summary":"随机森林填补缺失值(实测回归比较好) 1 2 3 4 5 %matplotlib inline from sklearn.tree import DecisionTreeClassifier from sklearn.ensemble import RandomForestClassifier#随机森林分类树 from sklearn.datasets import load_wine from sklearn.model_selection import train_test_split 1 2 wine = load_wine() wine …","tags":["python","数学建模","神经网络","机器学习"],"title":"预处理2-随机森林填补缺失值","url":"https://www.spiritysdx.top/20210723/"},{"categories":["数学建模","python","机器学习"],"date":"2021-07-22T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210722/index.md","section":"posts","summary":"前言 使用的相关数据链接如下： 点我跳转 数据预处理 无量纲化 1 2 #归一化 收到0~1之间 #正则化不是归一化 1 2 3 4 5 6 7 8 from sklearn.preprocessing import MinMaxScaler data = [[-1, 2], [-0.5, 6], [0, 10], [1, 18]] #不太熟悉numpy的小伙伴，能够判断data的结构吗？ #如果换成表是什么样子？ import pandas as pd pd.DataFrame(data) ","tags":["python","数学建模","神经网络","机器学习"],"title":"预处理1-分类数据","url":"https://www.spiritysdx.top/20210722/"},{"categories":["数学建模","python","机器学习"],"date":"2021-07-21T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210721/index.md","section":"posts","summary":"前言 数据集来自kaggle 链接：https://www.kaggle.com/c/titanic/data 里面的test和train的csv数据集为所需数据集。 ","tags":["python","数学建模","神经网络","机器学习"],"title":"决策树3-泰坦尼克号实例(网格搜索,超参数调参)","url":"https://www.spiritysdx.top/20210721/"},{"categories":["数学建模","python","机器学习"],"date":"2021-07-20T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210720/index.md","section":"posts","summary":"代码 1 2 3 from sklearn import tree from sklearn.datasets import load_wine from sklearn.model_selection import train_test_split 1 wine = load_wine() 1 wine.data.shape (178, 13) 1 wine.target array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, …","tags":["python","数学建模","神经网络","机器学习"],"title":"决策树2-简单调参","url":"https://www.spiritysdx.top/20210720/"},{"categories":["数学建模","python","机器学习"],"date":"2021-07-19T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210719/index.md","section":"posts","summary":"代码 1 2 3 4 5 import numpy as np import matplotlib.pyplot as plt from sklearn import datasets from sklearn.model_selection import train_test_split #分割训练集和测试集 from sklearn.neighbors import KNeighborsClassifier #K近邻 简单案例 1 2 3 iris=datasets.load_iris() iris_X = iris.data …","tags":["python","数学建模","神经网络","机器学习"],"title":"决策树1-分类器","url":"https://www.spiritysdx.top/20210719/"},{"categories":["数学建模","python","机器学习"],"date":"2021-07-18T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210718/index.md","section":"posts","summary":"代码 1 2 3 import numpy as np from sklearn.tree import DecisionTreeRegressor #回归树 import matplotlib.pyplot as plt 1 2 3 #生成噪声曲线 rng = np.random.RandomState(1) rng.rand(2,3)#生成区分行列的二维数据，一维数据不分行列，生成0~1之间的数 array([[4.17022005e-01, 7.20324493e-01, 1.14374817e-04], [3.02332573e-01, …","tags":["python","数学建模","神经网络","机器学习"],"title":"决策树线性回归","url":"https://www.spiritysdx.top/20210718/"},{"categories":["数学建模","python"],"date":"2021-07-17T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210717/index.md","section":"posts","summary":"我的前提条件，实际能使用jupyternotebook都可 Ubuntu 20.0系统 miniconda3和jupyter 直接通过jupyter notebook安装库 代码： 查找解释器位置 1 2 import os os.sys.executable 示例结果： 1 \u0026#39;/usr/bin/python3\u0026#39; 安装示例模板 1 2 3 4 5 6 7 8 9 10 11 import os libs = { …","tags":["python","linux"],"title":"直接通过jupyternotebook安装库","url":"https://www.spiritysdx.top/20210717/"},{"categories":["python","数学建模"],"date":"2021-07-15T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210715/index.md","section":"posts","summary":"新型冠状病毒（COVID-19/2019-nCoV）疫情分析 源文档详见：博客相关资源-新冠疫情数据分析文件 重要说明 分析文档：完成度：代码质量 3:5:2 其中分析文档是指你数据分析的过程中，对各问题分析的思路、对结果的解释、说明(要求言简意赅，不要为写而写) ","tags":["数学建模","python"],"title":"2020新型冠状病毒（COVID-19/2019-nCoV）疫情分析(补档)","url":"https://www.spiritysdx.top/20210715/"},{"categories":["python","数学建模"],"date":"2021-07-14T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210714/index.md","section":"posts","summary":"前言 帮舍友整的，不知道具体实际意义。 代码 1 2 3 4 5 import numpy as np import pandas as pd data = pd.read_excel(r\u0026#39;C:\\Users\\祈LHL\\Desktop\\data.xlsx\u0026#39;) 1 data.head() cellnumber x-coordinate y-coordinate z-coordinate density z-velocity relative-z-velocity x-coordinate.1 y-coordinate z-face-area …","tags":["数学建模","python"],"title":"某区域流体数据处理(只会写操作，不知道具体意义)","url":"https://www.spiritysdx.top/20210714/"},{"categories":["数学建模"],"date":"2021-07-13T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210713/index.md","section":"posts","summary":"前言 感谢组员书写书面报告，代码部分由我书写，我写的很烂，将就看吧。 第一届长三角数学建模 B题 锅炉水冷壁温度曲线 1Stopt多元拟合 个人体会： 只能确定2个自变量1个因变量的拟合函数形式，更高维的无法寻找公式进行拟合。 ","tags":["matlab","数学建模","皮尔逊相关系数","熵值法","多元函数拟合"],"title":"多元回归和熵值评价法及多目标遗传优化算法","url":"https://www.spiritysdx.top/20210713/"},{"categories":["数学建模"],"date":"2021-07-09T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210709/index.md","section":"posts","summary":"前言 感谢组员的共同协作，做组长的有些东西帮不上实在抱歉。 改进的SIR差分模型 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 …","tags":["matlab","数学建模","python"],"title":"改进的SIR差分模型及三个模型的应用","url":"https://www.spiritysdx.top/20210709/"},{"categories":["电脑技巧"],"date":"2021-07-06T00:00:00+00:00","lastmod":"2023-10-22T12:44:14+08:00","markdown_url":"https://www.spiritysdx.top/20210706/index.md","section":"posts","summary":"一、使用说明 更新一个整库脚本 1 ql repo \u0026lt;repourl\u0026gt; \u0026lt;path\u0026gt; \u0026lt;blacklist\u0026gt; \u0026lt;dependence\u0026gt; \u0026lt;branch\u0026gt; 更新单个脚本文件 1 ql raw \u0026lt;fileurl\u0026gt; 二、拉取整库实例【以下仓库排名不分先后，纯粹随机排列】 （一）某已退圈并不愿透露姓名大佬库 的现存备份托管库`（下列方案选1个就行，排名不分先后） ","tags":["linux"],"title":"JD脚本仓库合集","url":"https://www.spiritysdx.top/20210706/"},{"categories":["数学建模"],"date":"2021-07-05T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210705/index.md","section":"posts","summary":"7个组长分60个上机座位 问题 1.5月3日从早晨到晚上上机，207机房有60个机位, 请7个组长分上机座位 2.按照ABCDE组顺序开始分配, 每个组长都是足够聪明,A组长制定分配方案时,剩余的组长投票, 如有一半人不同意, A组就失去上机机会,同时失 去再次分配的权利,A组也失去投票机会. ","tags":["博弈论","数学建模"],"title":"海盗分金问题变种-组长分座位","url":"https://www.spiritysdx.top/20210705/"},{"categories":["电脑技巧"],"date":"2021-07-05T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210707/index.md","section":"posts","summary":"共用提取码 1 QLHL 蓝奏云盘 电脑Telegram 远程SSH工具,下载解压即可使用 各大文库PDF下载神器冰点文库 ","tags":["windows","实用软件"],"title":"实用软件","url":"https://www.spiritysdx.top/20210707/"},{"categories":["电脑技巧"],"date":"2021-07-04T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210704/index.md","section":"posts","summary":"CentOS8安装VNC窗口桌面并使用 Step1 安装vnc软件 终端窗口输入 1 2 sudo -i dnf install tigervnc tigervnc-server Step2 修改vncserver-config-defaults，如果添加一行localhost，外部不能访问 终端窗口输入 ","tags":["VNC","linux"],"title":"CentOS8的VNC窗口桌面使用","url":"https://www.spiritysdx.top/20210704/"},{"categories":["博客建站相关"],"date":"2021-07-03T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210703/index.md","section":"posts","summary":"鸣谢： 中文主题配置 主题文档 - 基本概念 探索 Hugo - LoveIt 主题的全部内容和背后的核心概念. ","tags":["hugo"],"title":"hugo的Lovelt主题配置","url":"https://www.spiritysdx.top/20210703/"},{"categories":["电脑技巧"],"date":"2021-07-02T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210702/index.md","section":"posts","summary":"官方安装教程 官方教程 个人使用经验 压缩包里的是适配Linux环境的二进制文件 按照官方文档操作后，发现一些小问题 首先是config.yml的部分配置应该是这样子的 ","tags":["linux"],"title":"go-cqhttp搭建教程","url":"https://www.spiritysdx.top/20210702/"},{"categories":["电脑技巧"],"date":"2021-07-01T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210701/index.md","section":"posts","summary":"2.8版本青龙面板搭建教程 step1 需要有一个云服务器，推荐我自用的腾讯云无忧计划： 优惠渠道 优点：每月续费同价，可随时重置服务器系统。 我的配置： 1 2 3 4 5 1核2g内存的北京-轻量云服务器 每月续费15元 系统应用重置为Ubuntu20.0版本 学生机优惠 ","tags":["linux"],"title":"青龙面板搭建教程","url":"https://www.spiritysdx.top/20210701/"},{"categories":["数学建模"],"date":"2021-06-27T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210627/index.md","section":"posts","summary":"前言 感谢组员的共同协作。 问题 以人口总数、人口年度增长率为研究对象，利用并根据世界人口网搜索1959年到2018年印度国家人口统计数据进行参数估计，即数据拟合，并对2019年印度人口进行增长预测。 用数学建模预测人口增长的方法主要有差分方程、微分方程、回归分析、时间序列等，结合题目、搜索到的数据以及《常微分方程》课本中所学知识，本小组以微分方程形式表示的改进指数增长模型、logistic模型为基础，以时间序列模型为拓展课题，建立以时间为自变量的印度人口增长模型。利用历史数据带入模型求解并做出预测。 ","tags":["matlab","数学建模"],"title":"基于常微分及时间序列模型的印度人口增长预测","url":"https://www.spiritysdx.top/20210627/"},{"categories":["C++","数据结构"],"date":"2021-06-25T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210625/index.md","section":"posts","summary":"前言 感谢组员的共同协作。 题目背景 迷宫问题是取自心理学的一个古典实验。在该实验中，将一只老鼠放入一个无顶大盒子的门口处，在出口处放置一块奶酪，奶酪吸引老鼠在盒子中寻找出口。对同一只老鼠进行反复实验，最终老鼠学会走通迷宫路线并不走错一步。 ","tags":["C++","数据结构"],"title":"多重算法实现的迷宫求解问题","url":"https://www.spiritysdx.top/20210625/"},{"categories":["C++"],"date":"2021-06-20T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210620/index.md","section":"posts","summary":"前言 感谢组员的共同协作。 四阶龙格-库塔公式计算微分方程数值解 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 #include\u0026lt;iostream\u0026gt; #include\u0026lt;iomanip\u0026gt; using namespace std; double …","tags":["C++","数值逼近"],"title":"几种数值算法的简单分析","url":"https://www.spiritysdx.top/20210620/"},{"categories":["C++","数据结构"],"date":"2021-05-19T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210519/index.md","section":"posts","summary":"题目 中序线索二叉链表的建立及遍历(数据结构严蔚敏C语言版的C++实现) 输入：字符串序列 输出：结点的相关信息，中序序列 处理方法: 1 2 3 4 5 6 1)在中序遍历过程中修改结点的左、右指针域，以保存当前访问结点的“前驱”和“后继”信息。 2)遍历过程中，附设指针pre, 并始终保持指针pre指向当前访问的指针p所指结点的前驱。 3)中序线索二叉树结构对称。其中：第一个结点是最左下的结点，最后一个结点是最右下的结点。 4)在中序线索二叉树上找结点的(直接)后继/前驱方法： a)若该结点有右孩子，其后继为其右子树中最左下的结点； b)若该结点无右 …","tags":["C++","数据结构"],"title":"C++中序线索二叉链表的建立及遍历","url":"https://www.spiritysdx.top/20210519/"},{"categories":["C++","数据结构"],"date":"2021-05-17T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210517/index.md","section":"posts","summary":"题目 统计二叉树中叶子结点的个数，计算二叉树的深度(数据结构严蔚敏C语言版的C++实现) 输入：字符串序列 输出：叶子结点的个数，二叉树的深度 处理方法： 1 2 1)先序遍历二叉树。在遍历过程中查找叶子结点，并计数。由此，需在遍历算法中增添一个“计数”的参数，并将算法中“访问结点”的操作改为：若是叶子，则计数器增1。 2)后序遍历二叉树。从二叉树深度的定义可知，二叉树的深度应为其左、右子树深度的最大值加1。由此，先分别求得左、右子树的深度，算法中“访问结点”的操作为：求得左、右子树深度的最大值，然后加 1 。 一、问题分析 统计叶子结点的个数需要先序遍 …","tags":["C++","数据结构"],"title":"C++二叉树中叶子结点的个数与深度的统计","url":"https://www.spiritysdx.top/20210517/"},{"categories":["C++","数据结构"],"date":"2021-05-16T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210516/index.md","section":"posts","summary":"题目 二叉链表的建立，先（中、后）序遍历(数据结构严蔚敏C语言版的C++实现) 输入：字符串序列 输出：先（中、后）序序列 处理方法：通过补虚结点，使二叉树中各实际结点均具有左右孩子，再对该二叉树按先序遍历进行输入。以字符串的形式:根、左子树、右子树定义一棵二叉树： ","tags":["C++","数据结构"],"title":"C++二叉树的建立","url":"https://www.spiritysdx.top/20210516/"},{"categories":["C++"],"date":"2021-05-15T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210515/index.md","section":"posts","summary":"数值逼近 公元 1225 年，比萨的数学家 Leonardo（即 Fibonacci（斐波那契）），1170-1250）研究了方程 x^3+2*x^2+10*x-20=0 得到一个根 x* = 1.368 808 107，没有人知道他用什么方法得到这个值。对于这个方程，分别用下列方法： （1）迭代法式1 ； （2）迭代法式2 ； （3）对（1）的 Aitken 加速方法； （4）对（2）的 Aitken 加速方法； （5）Newton 法 。 求方程的根（可取 x0 = 1 ），计算到 Leonardo 所得到的准确度。 ","tags":["C++","数值逼近"],"title":"C++数值逼近-迭代法Aikten法以及牛顿法求解线性方程根通用程序","url":"https://www.spiritysdx.top/20210515/"},{"categories":["python","数学建模"],"date":"2021-05-12T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210512/index.md","section":"posts","summary":"新型冠状病毒（COVID-19/2019-nCoV）疫情分析 spiritLHL 重要说明 帮同一个选修课的学妹码的结课作业，这是我个人完善后的版本(她的还有很多错漏) 分析文档：完成度：代码质量 3:5:2 其中分析文档是指你数据分析的过程中，对各问题分析的思路、对结果的解释、说明(要求言简意赅，不要为写而写) ","tags":["python","数学建模"],"title":"2021新型冠状病毒（COVID-19/2019-nCoV）疫情分析","url":"https://www.spiritysdx.top/20210512/"},{"categories":["C++"],"date":"2021-04-26T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210426/index.md","section":"posts","summary":"数值逼近 用迭代法求非线性方程(x+1)^2-sinx-3=0的根。 取迭代函数0.5*(3+sin((x))-1-(x)*(x))，精度要求为1*10^-2，最多迭代 100 次。 ","tags":["C++","数值逼近"],"title":"C++数值逼近-迭代法求非线性方程的根","url":"https://www.spiritysdx.top/20210426/"},{"categories":["C++"],"date":"2021-04-25T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210425/index.md","section":"posts","summary":"数值积分 取步长 h = 0.1，分别用 Euler 方法、改进 Euler 方法和经典四阶 Runge-Kutta 方法求解初值问题 1 2 f(x,y) = ((y)-2*(x)/(y)) f(0) = 0 并将计算结果与sqrt(1+2*(x))精确解相比较。 ","tags":["C++","数值逼近"],"title":"C++数值逼近-Euler方法、改进Euler方法和经典四阶Runge-Kutta方法求解初值问题","url":"https://www.spiritysdx.top/20210425/"},{"categories":["C++"],"date":"2021-04-24T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210424/index.md","section":"posts","summary":"数值逼近 将区间[0,1]四等分，分别用复化梯形公式和复化 Simpson 公式计算定积分 1 f(x) 4.0 / (1 + (x) * (x)) 的近似值。 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 #include\u0026lt;iostream\u0026gt; using namespace std; #include\u0026lt;stdio.h\u0026gt; //显示中文 …","tags":["C++","数值逼近"],"title":"C++数值逼近-复化梯形公式和复化 Simpson 公式计算定积分的近似值","url":"https://www.spiritysdx.top/20210424/"},{"categories":["C++"],"date":"2021-04-23T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210423/index.md","section":"posts","summary":"数值逼近 天安门广场升旗的时间是日出的时刻，而降旗的时间是日落时分。根据天安门广场管理委员会的公告，某年 10 月份升降旗的时间如下： 1 2 3 日期 1 15 22 升旗 6:09 6:23 6:31 降旗 17:58 17:36 17:26 请根据上述数据构造 Newton 插值多项式，并计算当年 10 月 8 日北京市的日照时长。 ","tags":["C++","数值逼近"],"title":"C++数值逼近-Newton插值法","url":"https://www.spiritysdx.top/20210423/"},{"categories":["C++"],"date":"2021-04-22T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210422/index.md","section":"posts","summary":"数值逼近 参考matlab程序：https://www.ilovematlab.cn/thread-450391-1-1.html 个人编写的c++版本如下 ","tags":["C++","数值逼近"],"title":"C++数值逼近-Hermite插值法通用程序","url":"https://www.spiritysdx.top/20210422/"},{"categories":["C++","数据结构"],"date":"2021-04-17T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210417/index.md","section":"posts","summary":"题目 行编辑(数据结构严蔚敏C语言版的C++实现) 输入：一行有误的数据 输出：一行正确的数据 处理方法： 允许用户输入出差错，并在发现有误时及时更正。例如：可用一个退格符“#”表示前一个字符无效；可用一个退行符“@”表示当前行中的字符均无效。 ","tags":["C++","数据结构"],"title":"C++栈结构实现行编辑","url":"https://www.spiritysdx.top/20210417/"},{"categories":["C++","数据结构"],"date":"2021-04-15T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210415/index.md","section":"posts","summary":"题目 括号匹配检验(数据结构严蔚敏C语言版的C++实现) 输入：一串括号 输出：检验结果（缺左括号、有多余左括号、左右不匹配、匹配成功） 处理方法： 检验括号是否匹配的方法可用“期待的急迫程度”这个概念来描述： ","tags":["C++","数据结构"],"title":"C++栈结构实现括号匹配","url":"https://www.spiritysdx.top/20210415/"},{"categories":["C++","数据结构"],"date":"2021-04-14T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210414/index.md","section":"posts","summary":"题目 数制转换(数据结构严蔚敏C语言版的C++实现) 输入：十进制整数 输出：八进制整数 处理方法：基于下列原理： 1 N = (N div d)×d + N mod d (其中：div为整除运算，mod为求余运算) ","tags":["C++","数据结构"],"title":"C++栈结构实现进制转换","url":"https://www.spiritysdx.top/20210414/"},{"categories":["C++","数据结构"],"date":"2021-04-13T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210413/index.md","section":"posts","summary":"题目 尾插法(数据结构严蔚敏C语言版的C++实现) 输入：键盘输入 输出：带头结点的单链表L 处理方法：待插结点*s插入到最后一个结点之后。 步骤： 1 2 3 4 5 1)获得最后一个结点的位置,使p指向该结点 2)p-\u0026gt;next = new LNode; 3)p = p-\u0026gt;next; 4)cin\u0026gt;\u0026gt;p-\u0026gt;data; 5)p-\u0026gt;next = NULL; 分析：要想获取最后一个结点的位置，必须从头指针开始顺着next链搜索链表的全部结点，该过程的时间复杂度是O( )。如果每次插入都按此方法获取最后一个结点的位置，则整个创建算法的时间复杂度为O( )。 ","tags":["C++","数据结构"],"title":"C++单链表尾插法","url":"https://www.spiritysdx.top/20210413/"},{"categories":["C++","数据结构"],"date":"2021-04-12T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210412/index.md","section":"posts","summary":"题目 按值非递减有序排列表合并(数据结构严蔚敏C语言版的C++实现) 已知线性表La和Lb中的数据元素按值非递减有序排列，要求将La和Lb归并成一个新的线性表Lc，且Lc中的数据元素仍按值非递减有序排列。（用顺序表实现） ","tags":["C++","数据结构"],"title":"C++非递减有序顺序表合并","url":"https://www.spiritysdx.top/20210412/"},{"categories":["C++","数据结构"],"date":"2021-04-11T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210411/index.md","section":"posts","summary":"题目 C++顺序表合并(数据结构严蔚敏C语言版的C++实现) 求两个线性表La和Lb的并集La = La U Lb （用顺序表实现） ( 要求：“就地运算”，运算结果仍然存放在La中 ) ","tags":["C++","数据结构"],"title":"C++顺序表合并","url":"https://www.spiritysdx.top/20210411/"},{"categories":["C++"],"date":"2021-03-08T00:00:00+00:00","lastmod":"2023-10-22T12:44:14+08:00","markdown_url":"https://www.spiritysdx.top/20210308/index.md","section":"posts","summary":"vscode2021新版的C++函数分文件编写配置 如图所示，需要下载这个插件 1 C/C++ Project Generator 安装并启用后，使用方法如下 创建工程文件后目录结构如下 main.cpp文件使用自定义的swap函数，swap函数的头文件放include文件夹，源文件放src文件夹 ","tags":["C++"],"title":"C++函数分文件编写(VScode2021配置教程)","url":"https://www.spiritysdx.top/20210308/"},{"categories":["C++"],"date":"2021-02-06T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210206/index.md","section":"posts","summary":"指针的概念 内存地址 内存空间的编排是以字节为单位的，每一个字节（也称单元）都有一个编号，这就是地址。内存地址通常是一个大小为4个字节的整数（这与具体的计算机有关），一般用十六进制数表示。 ","tags":["C++"],"title":"C++指针","url":"https://www.spiritysdx.top/20210206/"},{"categories":["C++"],"date":"2021-02-02T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210202/index.md","section":"posts","summary":"结构体 C++里数组是由相同数据类型的一组元素构成的集合，但实际问题中，经常会遇到由相互关联的，但类型不同的数据组成的数据结构。比如：描述一个学生的数据实体应该包括其学号、姓名、性别、年龄、成绩等数据。如果用多个独立的简单数据类型表示的话，就无法体现其整体性，也不便于数据的整体操作。 ","tags":["C++"],"title":"C++自定义数据类型","url":"https://www.spiritysdx.top/20210202/"},{"categories":["matlab","数学建模"],"date":"2021-02-01T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210201/index.md","section":"posts","summary":"代码 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 …","tags":["matlab","数学建模"],"title":"小额贷款的划算问题(matlab数学建模)","url":"https://www.spiritysdx.top/20210201/"},{"categories":["C++"],"date":"2021-01-28T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210128/index.md","section":"posts","summary":"前言 这部分类同于Python，比如从0开始，都用[]来表示第几个元素。 一维数组(score) 形式如下 1 \u0026lt;类型标识符\u0026gt;\u0026lt;数组名\u0026gt;[数组长度] 实例 1 2 3 4 5 6 int score[10]; int const N = 10; int n,m; int a[N]; //合法，N为常量 double[n],c[m*2]; //非法，n和m是变量 如果要访问一维数组的元素，形式如下： ","tags":["C++"],"title":"C++数组","url":"https://www.spiritysdx.top/20210128/"},{"categories":["C++"],"date":"2021-01-25T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210125/index.md","section":"posts","summary":"定义函数 1 2 3 4 \u0026lt;类型标识符\u0026gt;\u0026lt;函数名\u0026gt;(形式参数表) { 语句序列 } 实例： 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 //求n! #include\u0026lt;iostream\u0026gt; using namespace std; int factorial(int m); //原型声明 int main() { int n; cout\u0026lt;\u0026lt;\u0026#34;Input n:\u0026#34;; cin\u0026gt;\u0026gt;n; cout\u0026lt;\u0026lt;n\u0026lt;\u0026lt;\u0026#34;!=\u0026#34;\u0026lt;\u0026lt;\u0026#34;factorial(n)\u0026#34;\u0026lt;\u0026lt;endl; //函数调用作为表达式出现在语句中 …","tags":["C++"],"title":"C++函数","url":"https://www.spiritysdx.top/20210125/"},{"categories":["电脑技巧"],"date":"2021-01-23T00:00:00+00:00","lastmod":"2023-10-22T12:44:14+08:00","markdown_url":"https://www.spiritysdx.top/20210123/index.md","section":"posts","summary":"C盘空间视图化显示软件 蓝奏云盘链接：点此跳转 具体操作教程 视频教程点击这里跳转 除了按照视频操作，如果你还发现什么占空间特别大的文件，欢迎评论区留言交流","tags":["windows"],"title":"C盘清理视图化操作","url":"https://www.spiritysdx.top/20210123/"},{"categories":["C++"],"date":"2021-01-19T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210119/index.md","section":"posts","summary":"输入信息控制循环 输入信息控制循环通常控制的是无限循环(循环次数不确定的循环)。 通常使用while(或do while)语句构建无限循环。 统计输入的字符个数 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 #include\u0026lt;iostream\u0026gt; using namespace std; int main() { char ch; int count=0; cout\u0026lt;\u0026lt;\u0026#34;输入字符串，以#结束：\u0026#34;; while(true) { cin\u0026gt;\u0026gt;ch; if(ch==\u0026#39;#\u0026#39;) break; cout++; } cout\u0026lt;\u0026lt;\u0026#34;共输入 …","tags":["C++"],"title":"C++循环结构补充","url":"https://www.spiritysdx.top/20210119/"},{"categories":["C++"],"date":"2021-01-18T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210118/index.md","section":"posts","summary":"选择控制语句 if else 这里不像Python，每个分支语句后不需要加:，直接加 (Tab)后写分支语句，条件判断需要小加括号括起来，Python中是空格加条件，这点也不同。 ","tags":["C++"],"title":"C++程序控制结构","url":"https://www.spiritysdx.top/20210118/"},{"categories":["C++"],"date":"2021-01-15T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210115/index.md","section":"posts","summary":"赋值表达式 常规类似python，同样有赋值运算的简写形式，如 1 2 3 b += 2 //b = b + 2 x *= y + 3 //x = x*(y+3) x += x -= x*x //x = x + (x = x - x*x) 逗号表达式 在C++中，逗号也是一个运算符，使用形式为 \u0026lt;表达式1\u0026gt;,\u0026lt;表达式2\u0026gt;,……,\u0026lt;表达式n\u0026gt; ","tags":["C++"],"title":"C++简单程序","url":"https://www.spiritysdx.top/20210115/"},{"categories":["C++"],"date":"2021-01-10T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20210110/index.md","section":"posts","summary":"如果有Python的基础，学习c++会感觉很别扭2333 如果是用VS写c++程序，注意打开文件夹包含.vscode文件夹，否则会运行找不到路径。 (Python就没那么麻烦，不用整json配置) ","tags":["C++"],"title":"C++入门","url":"https://www.spiritysdx.top/20210110/"},{"categories":["C++"],"date":"2021-01-09T00:00:00+00:00","lastmod":"2023-10-22T12:44:14+08:00","markdown_url":"https://www.spiritysdx.top/20210109/index.md","section":"posts","summary":"安装vscode 安装包 教程安装C++用到的配置 压缩包解压后放到你装c++程序的文件夹里，里面的配置文件需要根据教程的第五步改。 压缩包链接 ","tags":["C++"],"title":"C++环境安装","url":"https://www.spiritysdx.top/20210109/"},{"categories":["python","爬虫"],"date":"2020-08-31T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200831/index.md","section":"posts","summary":"创建并使用多线程 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 print(\u0026#39;主线程执行代码\u0026#39;) # 从 threading 库中导入Thread类 from threading import Thread from time import sleep # 定义一个函数，作为新线程执行的入口函数 def threadFunc(arg1,arg2): print(\u0026#39;子线程 开始\u0026#39;) print(f\u0026#39;线程函数参数是：{arg1}, …","tags":["python","爬虫"],"title":"创建新线程(通用)","url":"https://www.spiritysdx.top/20200831/"},{"categories":["电脑技巧"],"date":"2020-07-27T00:00:00+00:00","lastmod":"2023-10-22T12:44:14+08:00","markdown_url":"https://www.spiritysdx.top/20200727/index.md","section":"posts","summary":"第三方词库资源 密码：QLHL 第三方词库下载链接 安装及使用视频 B站视频 ","tags":["windows"],"title":"微软输入法进化类搜狗输入法","url":"https://www.spiritysdx.top/20200727/"},{"categories":["matlab","数学建模"],"date":"2020-07-02T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200702/index.md","section":"posts","summary":"通过下标引用矩阵元素 A(3,2)表示A矩阵第3行第2列的元素。 如若超出限制行列维数，自动扩展，未赋值的默认为0 通过序号来引用矩阵元素 A(3)等同于A(1,2) A(i,j)的序号为(j-1)×m+i ","tags":["matlab","数学建模"],"title":"Matlab02(矩阵运算)","url":"https://www.spiritysdx.top/20200702/"},{"categories":["matlab","数学建模"],"date":"2020-07-01T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200701/index.md","section":"posts","summary":"基础命令 1.打开文件夹 命令行窗口输入 cd 文件夹名 这里推荐先在文件管理器先创建后打开 2.赋值变量会在工作区显示 可在命令行窗口输入whos 或 who 可以查看变量属性和具体参数 ","tags":["matlab","数学建模"],"title":"Matlab01(基础命令)","url":"https://www.spiritysdx.top/20200701/"},{"categories":["python","数学建模"],"date":"2020-06-22T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200622/index.md","section":"posts","summary":"前言 本篇鸣谢 马川-燕大 的增删整理，王圣元 ——原创文章，与原文不同之处包含我的学习记录。 匹配Jupyter Notebook的ipynb文档链接下载地址在资源页面里 ","tags":["python","数学建模"],"title":"Matplotlib","url":"https://www.spiritysdx.top/20200622/"},{"categories":["python","数学建模"],"date":"2020-06-22T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200602/index.md","section":"posts","summary":"前言 本篇鸣谢 马川-燕大 的增删整理，王圣元 ——原创文章，与原文不同之处包含我的学习记录。 匹配Jupyter Notebook的ipynb文档链接下载地址在资源页面里 ","tags":["python","数学建模"],"title":"Pandas (下)","url":"https://www.spiritysdx.top/20200602/"},{"categories":["数据库"],"date":"2020-06-15T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200615/index.md","section":"posts","summary":"分组函数 功能：作统计使用，又称为聚合函数或统计函数或组函数 分类： sum 求和 avg 平均值 max 最大值 min 最小值 count 计算个数 特点： 1.sum，avg一般用于处理数值型，max，min，count可以处理任何类型 ","tags":["MySQL"],"title":"MySQL数据库04(常见函数)","url":"https://www.spiritysdx.top/20200615/"},{"categories":["python","数学建模"],"date":"2020-06-02T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200526/index.md","section":"posts","summary":"前言 本篇鸣谢 马川-燕大 的增删整理，王圣元 ——原创文章，与原文不同之处包含我的学习记录。 匹配Jupyter Notebook的ipynb文档链接下载地址如下 ","tags":["python","数学建模"],"title":"Pandas (上)","url":"https://www.spiritysdx.top/20200526/"},{"categories":["数据库"],"date":"2020-05-31T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200531/index.md","section":"posts","summary":"函数 概念:类似于java的方法，将-组逻辑语句封装在方法体中，对外暴露方法名 好处: 1、隐藏了实现细节 2、提高代码的重用性 调用: select 函数名(实参列表) [ from 表] ; ","tags":["MySQL"],"title":"MySQL数据库03(常见函数)","url":"https://www.spiritysdx.top/20200531/"},{"categories":["数据库"],"date":"2020-05-30T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200530/index.md","section":"posts","summary":"前言 图形界面客户端下载链接 https://spiritlhl.lanzous.com/id5vuji 客户端安装 1.配置证书 2.新建用户配置 3.启用 4.询问(navaicate)窗口是命令行窗口 5.ctrl+s保存指令 6.ctrl+鼠标滚动轴调整字体大小 7.指令末尾加分号 ","tags":["MySQL"],"title":"MySQL数据库02(图形界面客户端基本运行指令)","url":"https://www.spiritysdx.top/20200530/"},{"categories":["数据库"],"date":"2020-05-29T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200529/index.md","section":"posts","summary":"前言 MySQL数据库下载链接(5.5版本) https://spiritlhl.lanzous.com/id5vt2f 常用指令 1.查看当前所有的数据库 show dat abases; 2.打开指定的库 use 库名; 3.查看当前库的所有表 show_ tables; 4.查看其它库的所有表 show tables from 库名; 5.创建表 ","tags":["MySQL"],"title":"MySQL数据库01(cmder基本运行指令)","url":"https://www.spiritysdx.top/20200529/"},{"categories":["python","爬虫"],"date":"2020-05-25T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200525/index.md","section":"posts","summary":"前言 最好有Selenium的Web自动化的实际经验 运行基础：client库(0.52版本)，Appium Server，安卓SDK(含JDK环境)，USB调试模式下的手机(开发者模式) ","tags":["python","爬虫"],"title":"Appium自动化操作03(界面操作和adb命令)","url":"https://www.spiritysdx.top/20200525/"},{"categories":["python","爬虫"],"date":"2020-05-24T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200527/index.md","section":"posts","summary":"前言 最好有Selenium的Web自动化的实际经验 运行基础：client库(0.52版本)，Appium Server，安卓SDK(含JDK环境)，USB调试模式下的手机(开发者模式) ","tags":["python","爬虫"],"title":"Appium自动化操作02(元素定位及查看工具)","url":"https://www.spiritysdx.top/20200527/"},{"categories":["python","爬虫"],"date":"2020-05-23T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200523/index.md","section":"posts","summary":"前言 最好有Selenium的Web自动化的实际经验 本篇用到的相关软件链接： 链接: https://pan.baidu.com/s/126x-AgLKvM7qSJqdOzAAHA 提取码: h9b2 Appium 基础知识 Appium 用途和特点 Appium 是一个移动 App （手机应用）自动化工具。 ","tags":["python","爬虫"],"title":"Appium自动化操作01(环境安装与初始结构)","url":"https://www.spiritysdx.top/20200523/"},{"categories":["python","数学建模"],"date":"2020-05-22T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200522/index.md","section":"posts","summary":"前言 本篇鸣谢 马川-燕大 的增删整理，王圣元 ——原创文章，与原文不同之处包含我的学习记录。 匹配Jupyter Notebook的ipynb文档链接下载地址如下 ","tags":["python","数学建模"],"title":"NumPy (下)","url":"https://www.spiritysdx.top/20200522/"},{"categories":["python","爬虫"],"date":"2020-05-17T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200517/index.md","section":"posts","summary":"前言 需要下载Chrome或Firefox的driver，Chrome内核81.440与Firefox内核74.0下载链接如下： Firefox Chrome 其他版本请在搜索引擎查找，本篇使用该版本，注意，driver下载后需要配置对应内核的游览器，电脑本身需要有该内核的游览器。 ","tags":["python","爬虫"],"title":"Selenium的web自动化操作03(语法补充)","url":"https://www.spiritysdx.top/20200517/"},{"categories":["python","数学建模"],"date":"2020-05-12T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200512/index.md","section":"posts","summary":"前言 本篇鸣谢 马川——燕大 增删整理，王圣元——原创文章，与原文不同之处包含我的学习记录。 匹配Jupyter Notebook的ipynb文档链接下载地址如下 ","tags":["python","数学建模"],"title":"NumPy (上)","url":"https://www.spiritysdx.top/20200512/"},{"categories":["python","爬虫"],"date":"2020-05-10T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200510/index.md","section":"posts","summary":"前言 需要下载Chrome或Firefox的driver，Chrome内核81.440与Firefox内核74.0下载链接如下： Firefox Chrome 其他版本请在搜索引擎查找，本篇使用该版本，注意，driver下载后需要配置对应内核的游览器，电脑本身需要有该内核的游览器。 ","tags":["python","爬虫"],"title":"Selenium的web自动化操作01(环境布置与标准流程)","url":"https://www.spiritysdx.top/20200510/"},{"categories":["python","数学建模"],"date":"2020-05-09T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200509/index.md","section":"posts","summary":"前言 本篇鸣谢 燕大——马川 的整理 匹配Jupyter Notebook的ipynb文档链接下载地址如下 源文档 Python编程基础 by 马川 燕大 代码胜于雄辩 Talks is cheap. Show me the code. —-Linus Torvalds(Linux操作系统的奠基者) ","tags":["python","数学建模"],"title":"Python数据分析快速入门","url":"https://www.spiritysdx.top/20200509/"},{"categories":["python"],"date":"2020-05-05T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200505/index.md","section":"posts","summary":"直接放代码 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 from selenium import webdriver import time driver = webdriver.Chrome(r\u0026#39;C:\\chromedriver.exe\u0026#39;) urllist = [ …","tags":["python"],"title":"比较简单的selenium自动化操作播放bilibili(b站)视频2020","url":"https://www.spiritysdx.top/20200505/"},{"categories":["python","爬虫"],"date":"2020-05-05T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200506/index.md","section":"posts","summary":"前言 re库的实用实例如下 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 import requests import re import os a = True while a: #创建一个文件夹，保存所有图片 if not os.path.exists(\u0026#39;./tupianLibs\u0026#39;): os.mkdir(\u0026#39;./tupianLibs\u0026#39;) headers = { \u0026#39;User-Agent\u0026#39;:\u0026#39;Mozilla/5.0 …","tags":["python","爬虫"],"title":"爬虫流程及方法14(正则表达式篇)","url":"https://www.spiritysdx.top/20200506/"},{"categories":["python","爬虫"],"date":"2020-05-04T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200504/index.md","section":"posts","summary":"前言 xpath解析原理: 1.实例化一个etree的对象，且需要将被解析的页面源码数据加载到该对象中。 2.调用et ree对象中的xpath方法结合着xpath表达式实现标签的定位和内容的捕获。 ","tags":["python","爬虫"],"title":"爬虫流程及方法13(xpath解析页面)","url":"https://www.spiritysdx.top/20200504/"},{"categories":["python","爬虫"],"date":"2020-05-01T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200501/index.md","section":"posts","summary":"前言 目的:在爬虫中使用异步实现高性能的数据爬取操作。 异步爬虫的方式: –多线程，多进程(不建议): 好处:可以为相关阻塞的操作单独开启线程或者进程，阻塞操作就可以异步执行。 弊端:无法无限制的开启多线程或者多进程。 ","tags":["python","爬虫"],"title":"爬虫流程及方法12(高性能异步爬虫)","url":"https://www.spiritysdx.top/20200501/"},{"categories":["python","爬虫"],"date":"2020-04-29T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200429/index.md","section":"posts","summary":"原网页链接萌新论坛 requests 伪装 headers 发送请求 headers中空着的可能有也可能无，user-agent基本得有 在chrome中找到网页的请求头，图片如下 ","tags":["python","爬虫"],"title":"爬虫流程及方法10(爬虫伪装专题篇)","url":"https://www.spiritysdx.top/20200429/"},{"categories":["python","爬虫"],"date":"2020-04-28T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200428/index.md","section":"posts","summary":"前言 本篇鸣谢 清华——尹成 的整理收集 PyQuery文档https://www.osgeo.cn/pyquery/index.html PyQuery库也是一个非常强大又灵活的网页解析库，如果你有前端开发经验的，都应该接触过jQuery,那么PyQuery就是你非常绝佳的选择，PyQuery 是 Python 仿照 jQuery 的严格实现。语法与 jQuery 几乎完全相同，所以不用再去费心去记一些奇怪的方法了。 官网地址:http://pyquery.readthedocs.io/en/latest/ jQuery参考文档 …","tags":["python","爬虫"],"title":"爬虫流程及方法11(PyQuery解析网页篇)(全)","url":"https://www.spiritysdx.top/20200428/"},{"categories":["python","爬虫"],"date":"2020-04-26T00:00:00+00:00","lastmod":"2023-10-22T12:44:14+08:00","markdown_url":"https://www.spiritysdx.top/20200426/index.md","section":"posts","summary":" 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 import requests from bs4 import BeautifulSoup a = True while a: headers = { \u0026#39;User-Agent\u0026#39;: \u0026#39;Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) …","tags":["python","爬虫"],"title":"爬虫流程及方法08(BeautifulSoup实例)(非ajax请求)","url":"https://www.spiritysdx.top/20200426/"},{"categories":["python","爬虫"],"date":"2020-04-26T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200427/index.md","section":"posts","summary":"动态加载数据ajax 首页中对应企业数据通过ajax请求得到 详情页url只有id不同其余相同 id从json中获取，域名与id拼接新url 详情页的数据也是动态加载出来的 详情页的url也是相同的只有id不同 ","tags":["python","爬虫"],"title":"爬虫流程及方法09(动态加载页面)(ajax请求)(Json实例)","url":"https://www.spiritysdx.top/20200427/"},{"categories":["博客建站相关"],"date":"2020-04-25T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200425/index.md","section":"posts","summary":"1. 前言 本markdown文档初版下载链接： (建议看初版文档理解书写格式，书写标准以本网站本页面为主) https://pan.baidu.com/s/1z_2IsuaRh8cYmtssepvIXQ 提取码：0a83 本篇鸣谢胡国磊学长整理，由我进行加工发布 ","tags":["hugo"],"title":"撰写hugo博客的markdown格式","url":"https://www.spiritysdx.top/20200425/"},{"categories":["python","爬虫"],"date":"2020-04-22T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200422/index.md","section":"posts","summary":"技术路线 1.requests-BeautifulSoup 2.scrapy(5+2结构) 3.scrapy + requests-Beautiful-re + PhantomJS —\u0026gt;表单提交、爬取周期、入库存储(js处理) 4.requests-xpath 5.requests-ccs 6.requests库可与urllib库互换 ","tags":["python","爬虫"],"title":"爬虫流程及方法07(爬虫技术路线整理)(实时更新)","url":"https://www.spiritysdx.top/20200422/"},{"categories":["python","爬虫"],"date":"2020-04-21T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200421/index.md","section":"posts","summary":"提高爬取速度的方法 1.在setting.py文件里修改并发选项 2.使用scrapy-*的高级补充库，特化某方面，提升速度 ","tags":["python","爬虫"],"title":"爬虫流程及方法06(Scrapy进阶爬虫)(实时更新)","url":"https://www.spiritysdx.top/20200421/"},{"categories":["python","爬虫"],"date":"2020-04-20T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200420/index.md","section":"posts","summary":"Request类 1 class scrapy.http.Request() Request对象表示一个HTTP请求。 由Spider生成，由Downloader执行。 常用属性： ","tags":["python","爬虫"],"title":"爬虫流程及方法05(Scrapy入门级爬虫)","url":"https://www.spiritysdx.top/20200420/"},{"categories":["python","爬虫"],"date":"2020-04-17T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200417/index.md","section":"posts","summary":"安装scrapy pycharm安装步骤: 1.打开左上角file 2.打开Other Setting下的Setting for New Project 3.在Project Interpreter选择Project Interpreter里你使用的编译器后，点击加号(+)添加包 ","tags":["python","爬虫"],"title":"爬虫流程及方法04(Scrapy框架)","url":"https://www.spiritysdx.top/20200417/"},{"categories":["电脑技巧"],"date":"2020-04-12T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200412/index.md","section":"posts","summary":"个人收集，网站有可能不安全 建议浏览器隐私模式下使用，避免用户信息被盗窃 该页面下载共享提取码： 1 QLHL 该页面解析地址由个人收集，替代网页插件版的直链解析，原理是pandownload的网页版，需要百度云盘的文件共享链接。 ","tags":["windows"],"title":"百度云盘加速","url":"https://www.spiritysdx.top/20200412/"},{"categories":["博客建站相关"],"date":"2020-04-11T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200411/index.md","section":"posts","summary":"本地搭建博客 创建新文章 hugo new 你的文档名/你的文章名.md 在码云中创建库 1.链接一定是：/(填你的用户名) 2.选择公开/私有都行 3.注意不要初始化库(三个选项都不要选) 4.创建库 5.复制库链接(https类型) ","tags":["hugo"],"title":"hugo博客部署码云","url":"https://www.spiritysdx.top/20200411/"},{"categories":["机器学习"],"date":"2020-04-09T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200409/index.md","section":"posts","summary":"教计算机识别手写数字 (转载自YouTube) 人工神经网络是在现代神经科学的基础上提出和发展起来的，旨在反映人脑结构及功能的一种抽象数学模型。自1943 年美国心理学家 W. McCulloch 和数学家 W. Pitts 提 出形式神经元的抽象数学模型—MP 模型以来，人工神经网络理论技术经过了 50 多年 曲折的发展。特别是 20 世纪 80 年代，人工神经网络的研究取得了重大进展，有关的理论和方法已经发展成一门界于物理学、数学、计算机科学和神经生物学之间的交叉学科。希望阅读本篇，能使你对神经网络有一个大概的了解，明白其大概的工作原理。 ","tags":["机器学习","深度学习","Python","神经网络"],"title":"深度学习之神经网络","url":"https://www.spiritysdx.top/20200409/"},{"categories":["博客建站相关"],"date":"2020-04-04T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200404/index.md","section":"posts","summary":"本地搭建博客 创建新文章 hugo new 你的文档名/你的文章名.md 创建库 1.名称一定是：你的用户名.github.io 2.选择本地存储复制 ","tags":["hugo"],"title":"hugo博客在GitHub上进行部署","url":"https://www.spiritysdx.top/20200404/"},{"categories":["python","爬虫"],"date":"2020-04-01T00:00:00+00:00","lastmod":"2023-10-22T12:44:14+08:00","markdown_url":"https://www.spiritysdx.top/20200401/index.md","section":"posts","summary":" 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 #!/usr/bin/env python3 #本篇介绍抓取含搜索引擎的爬虫 #UA检测：门户网站检测对应请求的身份标识 #UA：useragent（请求载体的身份标识） #UA伪装：伪装游览器 import requests a = True while a: # UA伪装：伪装游览器,将对应user-agent封装到字典headers中 headers = { …","tags":["python","爬虫"],"title":"爬虫流程及方法03(搜索引擎爬取)","url":"https://www.spiritysdx.top/20200401/"},{"categories":["python","爬虫"],"date":"2020-03-31T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200331/index.md","section":"posts","summary":" 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 #!/usr/bin/env python3 #对某论坛的爬取 import requests from bs4 import BeautifulSoup import time #需求：爬取网站标题及详情页的文本 a = True while a:#可转变成实时循环#对首页的页面数据进行爬取 url = …","tags":["python","爬虫"],"title":"爬虫流程及方法02(Beautiful Soup解析页面)","url":"https://www.spiritysdx.top/20200331/"},{"categories":["python","爬虫"],"date":"2020-03-20T00:00:00+00:00","lastmod":"2026-06-17T11:53:04+08:00","markdown_url":"https://www.spiritysdx.top/20200320/index.md","section":"posts","summary":"爬虫究竟是合法还是违法的? 在法律中是不被禁止 具有违法风险 请善意爬虫 切勿恶意爬虫 爬虫带来的风险可以体现在如下2方面: 爬虫干扰了被访问网站的正常运营 爬虫抓取了受到法律保护的特定类型的数据或信息 ","tags":["python","爬虫"],"title":"爬虫流程及方法01(入门准备及Request库使用)","url":"https://www.spiritysdx.top/20200320/"},{"categories":["python"],"date":"2020-03-15T00:00:00+00:00","lastmod":"2023-10-22T12:44:14+08:00","markdown_url":"https://www.spiritysdx.top/20200315/index.md","section":"posts","summary":"直接放代码 1 2 3 4 5 6 7 8 9 10 11 12 import os libs = { \u0026#34;requests\u0026#34;,\u0026#34;jieba\u0026#34;,\u0026#34;beautifulsoup4\u0026#34;,\\ \u0026#34;django\u0026#34;,\u0026#34;flask\u0026#34;,\\ \u0026#34;此处填写你需要下载的库的名称，注意大小写并拼写正确，样式如上面例子\u0026#34;,\u0026#34;pandas\u0026#34; } try: for lib in libs: os.system(\u0026#39;pip install \u0026#39;+lib) print(\u0026#34;Successful\u0026#34;) except: print(\u0026#39;error\u0026#39;) os.system(command) command 为要 …","tags":["python"],"title":"安装python第三方库的小技巧","url":"https://www.spiritysdx.top/20200315/"},{"categories":["博客建站相关"],"date":"2019-01-01T00:00:00+00:00","lastmod":"2023-10-22T12:44:14+08:00","markdown_url":"https://www.spiritysdx.top/20190101/index.md","section":"posts","summary":"创建新文档 1 hugo new posts/名字.md 引用B站视频 1 注意 如果要绑定第三方域名给github-pages，每次更新博客都需要重新绑定！ ","tags":["hugo"],"title":"本主题使用小技巧","url":"https://www.spiritysdx.top/20190101/"}],"site":"二叉树的博客","sitemap":"https://www.spiritysdx.top/sitemap.xml"}