3 Commits
Author SHA1 Message Date
treeandClaude Opus 5.5 7e690e7e6d 新增 CHANGELOG.md,补记 v0.1.0 ~ v0.3.0 版本更新记录
Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-20 20:16:48 +08:00
treeandClaude Opus 5.5 0e070a64fe P1:闭环时限管控、治理成效评估、站型阈值系数与接口测试
- 闭环状态流转自动记录接单/闭环时间,状态值校验;按断站分级判定响应时限
  (及时/超时/超时未响应/未到期/无记录),新增及时响应率、平均闭环时长、超时未闭环指标与筛选
- 新增治理成效评估 /api/effect:分界日期前后指标对比与问题站点逐站改善判定
- 新增站型阈值系数(实际阈值 = 场景阈值 × 站型系数)及配置接口与页面
- 样例数据新增接单/闭环时间(独立随机数,原字段不变),导入支持并校验
- 新增 tests/test_api.py(权限、导入模板、闭环流程、站型系数、成效评估、无数据边界)
- 修复:新流水无告警匹配时根因分析报错;空筛选时超长站点清单返回 500

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-18 15:14:06 +08:00
treeandClaude Opus 5.5 e9616af34d 修复 P0:根因判定、退服率口径、影响评估与告警关联量化
- 根因判定改为三级(衍生关系/时间窗/疑似)并带置信度;小区退服告警 ±2 分钟匹配;
  衍生告警按父告警归类;优先选随业务恢复清除的告警;时间窗按历史流水 95% 分位数标定
- 退服率按日历统计周期计算;站点数分母支持站名关键字;重复退服率除以全部站点;
  月均按统计周期自然月数;导出与看板同一口径
- 影响评估新增受影响用户·小时与 500 米邻站覆盖补偿
- 新增告警—退服关联量化(支持度/置信度/提升度)及 /api/association,根因页展示
- 新增 tests/(扰动数据集留出检验、指标口径检验)与 bench.py;旧库启动时自动迁移

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
2026-11-18 14:43:34 +08:00
19 changed files with 7463 additions and 6145 deletions

No files matched your search

+3
View File
@@ -18,3 +18,6 @@ Thumbs.db
desktop.ini desktop.ini
.DS_Store .DS_Store
~$* ~$*
# 测试生成的临时数据库与数据
tests/_work/
+71
View File
@@ -0,0 +1,71 @@
# 更新记录
版本号格式为 主版本.次版本.修订号。每个版本在仓库中有同名标签(`git show v0.2.0` 可查看说明)。
## v0.3.1
- 新增本更新记录 `CHANGELOG.md`,并为历史版本补打标签 v0.1.0 ~ v0.3.0。
- 仓库同时推送到 GitHub(treehell/gset)与 git.treemoe.cn(tree/Gset)。
## v0.3.0 — 闭环时限管控、治理成效评估、站型阈值系数
### 新增
- **闭环时限**:状态改为“处理中 / 已闭环”时自动记录接单 / 闭环时间,重新打开时清空闭环时间;
按断站分级判定响应时限(重要 2 小时、主要 4 小时、次要 24 小时),每条流水判定为
及时 / 超时 / 超时未响应 / 未到期 / 无记录。
- 影响评估页新增“闭环管控”:闭环率、及时响应率、平均闭环时长、超时未闭环条数;
退服流水管理新增“仅超时未闭环”筛选与响应时限列。
- **治理成效评估**(`/api/effect`):以分界日期把统计周期分成治理前后两段,各按天数折算后比较 8 项指标;
治理前的超长 / 高频站点逐站判定改善、持平或恶化(±10%)。
- **站型阈值系数**:站点实际阈值 = 场景阈值 × 站型系数(默认室分超长 ×0.75,分布式皮基站 ×1.5),
可在「场景与站点分级」页修改。
- 样例数据新增接单时间、闭环时间两列(独立随机数生成,原有字段不变);导入支持这两列并校验格式与先后顺序。
- `tests/test_api.py`:登录与权限、导入模板校验、闭环流程、站型系数、治理成效、无数据边界,共 41 项检查。
### 修复
- 新导入的流水没有任何告警能匹配时,根因分析报错,导致系统自带的导入模板必定导入失败(v0.2.0 引入)。
- 筛选结果为空时,超长站点清单接口返回 500。
- 闭环状态没有校验,普通用户可写入任意状态值;修改不存在的记录现在返回 404。
### 升级说明
- 旧数据库启动时自动补 `handled_at`、`closed_at` 字段和站型系数表;旧记录没有接单时间,及时响应率显示“—”。
删除 `data/outage.db` 后重启可载入带闭环时间的样例数据。
## v0.2.0 — 根因判定、退服率口径、影响评估、关联量化
### 修复
- **根因判定在非理想数据上准确率大幅下降。** 改为三级判定并带置信度:
1. 衍生关系(高):小区退服告警与退服开始相差 ≤ 2 分钟即匹配(原来要求时间完全相等);
2. 时间窗(中):时间窗按已闭环流水的提前量 95% 分位数标定,不再写死;
3. 疑似(低):时间窗放宽 3 倍命中,提示人工复核。
另外衍生告警按父告警归类,多条命中时优先选随业务恢复清除的告警。
留出集(7–9 月)准确率:
| 测试场景 | v0.1.0 | v0.2.0 |
|---|---|---|
| 仓库样例数据 | 98.42% | 99.28% |
| 小区退服告警时刻偏移 | 衍生判定 0% 命中 | 99.28% |
| 30% 干扰告警 | 89.30% | 96.08% |
| 告警提前量变宽 | 63.08% | 97.96% |
| 综合扰动 | 53.23% | 82.99% |
- **退服率**统计时长改为日历周期(原来取首末记录间隔,稀疏筛选时偏高达 78%)。
- 按站名关键字筛选时,站点数分母仍为全部站点(单站单日 185 → 1)。
- 导出报表与看板的退服率不一致(按场景筛选时相差 4–7 倍)。
- 单独查看某站时“月均”按该站有退服的月数计算,导致 23 个站超长 / 高频判定翻转。
- 重复退服率的分母改为全部站点。
### 新增
- 影响评估新增“受影响用户·小时”,并按 500 米内邻站数做覆盖补偿(原受影响用户与时长相关系数仅 0.03)。
- 告警—退服关联量化(`/api/association`):各类告警的支持度、置信度、提升度,根因页展示时间窗标定表与关联强度表。
- `tests/test_rootcause.py`(6 组扰动数据集留出检验)、`tests/test_kpi.py`、`bench.py`(性能与时长准确率)。
## v0.1.0 — 初始版本
- 用户权限(管理员 / 普通用户)、密码加盐哈希存储、操作日志。
- 退服流水、告警流、华为网管告警、5G 工参导入,逐行校验,预览确认后入库。
- 退服时长按起止时间重算,上报时长差错核对。
- 两级根因判定(衍生关系 + 固定时间窗)与分级处置建议。
- 场景阈值配置、超长 / 高频站点清单、影响评估、可视化看板、三类报表导出。
- 合成样例数据:南宁 7 个城区、185 个站、6000 条退服流水、5 类通信场景。
+35 -13
View File
@@ -25,10 +25,11 @@ Python + Flask + SQLite 后端,React 19 + [Liquid Glass UI](https://tsdsj.gith
``` ```
Local-sever/ Local-sever/
├── app.py Flask 路由:认证与权限、增删改查、导入、统计、导出、日志、用户;托管前端页面 ├── app.py Flask 路由:认证与权限、增删改查、导入、统计、导出、日志、用户;托管前端页面
├── core.py 核心算法:建库入库、两级根因关联、时长与考核指标 ├── core.py 核心算法:建库入库、时间窗标定、三级根因关联、告警关联量化、时长与考核指标
├── importer.py 数据导入:读取 Excel/CSV、自动识别文件类型、逐行校验(见下方“支持导入的文件”) ├── importer.py 数据导入:读取 Excel/CSV、自动识别文件类型、逐行校验(见下方“支持导入的文件”)
├── data_gen.py 第 18 题数据生成器(Data-Generated-Core-Py/data.py 的改进版,字段口径一致) ├── data_gen.py 第 18 题数据生成器(Data-Generated-Core-Py/data.py 的改进版,字段口径一致)
├── publish_online.py 同步到 ../Online-sever(1Panel 线上版) ├── bench.py 性能与时长准确率测试(对应开题报告预期目标)
├── tests/ 根因准确率扰动测试、考核指标口径测试(见下方“测试”)
├── data/ sites.csv / data_outage.csv / data_alarm.csv 种子数据,outage.db 自动生成 ├── data/ sites.csv / data_outage.csv / data_alarm.csv 种子数据,outage.db 自动生成
├── static/ 前端(全部可直接阅读和修改) ├── static/ 前端(全部可直接阅读和修改)
│ ├── index.html 页面入口,importmap 把 'vendor' / 'echarts' 指向 vendor/ 下的文件 │ ├── index.html 页面入口,importmap 把 'vendor' / 'echarts' 指向 vendor/ 下的文件
@@ -42,8 +43,6 @@ Local-sever/
│ │ ├── components/ ui(卡片/指标/标签/下拉)、DataTable(表格+分页)、FormDialog、EChart、icons │ │ ├── components/ ui(卡片/指标/标签/下拉)、DataTable(表格+分页)、FormDialog、EChart、icons
│ │ └── pages/ index.js 页面注册表 + 10 个业务页面,一页一个文件 │ │ └── pages/ index.js 页面注册表 + 10 个业务页面,一页一个文件
│ └── vendor/ 第三方库(React、液态玻璃组件库、htm、ECharts),压缩文件,不要手改 │ └── vendor/ 第三方库(React、液态玻璃组件库、htm、ECharts),压缩文件,不要手改
├── tools/vendor-build/ 只在升级第三方库时使用,见下文
└── _legacy_frontend/ 最早的原生 JS 前端,仅供对照,可删除
``` ```
## 修改前端 ## 修改前端
@@ -68,7 +67,7 @@ html`<${Section} title="标题" span=${6}><${EChart} option=${opt}/></${Section}
### 升级第三方库(很少需要) ### 升级第三方库(很少需要)
`static/vendor/` 由 `tools/vendor-build/` 一次性生成。只有升级 React、组件库或 ECharts,或需要新的 React API / 新的图表类型时才需要 Node.js 20.19+: `static/vendor/` 由 `tools/vendor-build/`(不在本仓库中)一次性生成。只有升级 React、组件库或 ECharts,或需要新的 React API / 新的图表类型时才需要 Node.js 20.19+:
```bash ```bash
cd tools/vendor-build cd tools/vendor-build
@@ -96,6 +95,8 @@ npm run build # 重新生成 ../../static/vendor/
重新生成数据:`python data_gen.py --rows 6000`,然后删除 `data/outage.db` 并重启。 重新生成数据:`python data_gen.py --rows 6000`,然后删除 `data/outage.db` 并重启。
样例流水含接单时间 `handle_time` 与闭环时间 `close_time`(独立随机数生成,不影响其他字段)。旧版数据库启动时会自动补字段,但旧记录没有这两个时间,及时响应率显示为“—”;删除 `data/outage.db` 重启即可载入带闭环时间的样例数据。
## 支持导入的文件(「数据导入校验」页,上传后自动识别) ## 支持导入的文件(「数据导入校验」页,上传后自动识别)
| 类型 | 文件 | 入库方式 | | 类型 | 文件 | 入库方式 |
@@ -112,12 +113,33 @@ npm run build # 重新生成 ../../static/vendor/
## 核心算法(core.py) ## 核心算法(core.py)
- **退服时长**:一律按“恢复时间 − 开始时间”重算;上报时长保存为 `reported_min`,两者相差超过 1 分钟即标记为“时长差错”。 - **退服时长**:一律按“恢复时间 − 开始时间”重算;上报时长保存为 `reported_min`,两者相差超过 1 分钟即标记为“时长差错”。
- **退服率** = 总退服时长 ÷(站点数 × 统计时长);**网络可用率** = 1 − 退服率。 - **统计周期**:顶栏筛选了月份 / 日期范围时取筛选范围(同时有则取交集),否则取全部流水首日 0 点到末日 24 点。
- **超长退服站点**:月均超长退服 ≥ 1 次(单次时长 ≥ 场景阈值)。**高频重复站点**:月均退服次数 ≥ 场景次数阈值。两个阈值都可在页面上配置。 - **退服率** = 总退服时长 ÷(站点数 × 统计周期);**网络可用率** = 1 − 退服率。站点数按场景、片区、站型、厂家、站名关键字筛选。看板、对标与导出用同一口径。
- **根因判定分两级**: - **站型分级阈值**:站点实际阈值 = 场景阈值 × 站型系数。默认室分超长时长 ×0.75(重点楼宇从严),分布式皮基站两项 ×1.5(覆盖小、有宏站兜底),宏站 ×1;可在「场景与站点分级」页修改。
1. 衍生关系:“小区退服”告警带有 `parent_alarm_id` 时,直接取父告警的类别。 - **超长退服站点**:月均超长退服 ≥ 1 次(单次时长 ≥ 站点实际阈值)。**高频重复站点**:月均退服次数 ≥ 站点实际次数阈值。“月均”按统计周期的自然月数计算。**重复退服率** = 高频重复站点数 ÷ 站点数。
2. 时间窗:回溯退服前未清除的同站告警,按优先级判定:割接施工(30 分钟)→ 电源故障(300 分钟)→ 机房环境(180 分钟)→ 传输中断(15 分钟)→ 天馈故障(30 分钟)→ 设备硬件(15 分钟)。 - **根因判定分三级**(每条结果带置信度 高 / 中 / 低):
0. 衍生告警归并:带 `parent_alarm_id` 的告警按父告警类别计(停电引起的传输告警归为电源故障)。
1. 衍生关系(高):与退服开始相差 ≤ 2 分钟的“小区退服”告警带父告警时,取父告警类别。
2. 时间窗(中):同站在退服时仍未清除、提前量在该类时间窗内的告警。
3. 疑似(低):时间窗放宽到 3 倍才命中,提示人工复核。
准确率按“上报原因大类 × 判定根因”交叉矩阵计算,样例数据约 98%。 多条命中时依次比较:级别 → 是否随业务恢复清除(±15 分钟)→ 规则优先级(割接施工 → 电源故障 → 机房环境 → 传输中断 → 天馈故障 → 设备硬件)→ 时间早;都没有命中判为“其他”。
- **断站分级**:重要 / 主要 / 次要,对应不同的处置响应时限。 - **时间窗标定**:每次全量分析前,用已闭环流水中同站同类告警的提前量 95% 分位数标定各类时间窗(不小于默认值 30/300/180/15/30/15 分钟,不大于 720 分钟;样本少于 30 条时沿用默认值)。
- **影响评估**:受影响用户 = 小区数 × 单小区在线用户 × 忙闲时系数;业务损失 = 小区话务 × 退服时长。 - **告警—退服关联量化**:对每类告警计算支持度(退服前 W 分钟内有该类告警的退服占比)、置信度(该类告警其后 W 分钟内发生退服的比例)、提升度(置信度 ÷ 随机窗口遇到退服的概率)。
- **断站分级**:重要 / 主要 / 次要,对应不同的处置响应时限(2 / 4 / 24 小时,从退服开始算到接单)。
- **闭环时限**:状态改为“处理中 / 已闭环”时自动记录接单时间,改为“已闭环”时自动记录闭环时间,重新打开时清空闭环时间。每条流水判定为 及时 / 超时 / 超时未响应 / 未到期 / 无记录(历史数据没有接单时间);看板给出及时响应率、平均闭环时长、超时未闭环条数,流水管理可筛选“仅超时未闭环”。
- **治理成效评估**:以分界日期把统计周期分为治理前后两段,各按自身天数折算后比较 8 项指标;治理前被判为超长 / 高频的站点逐站比较月均退服次数与时长,变化超过 ±10% 记为改善或恶化。退服有季节性,跨季节对比需结合同期数据判断。
- **影响评估**:受影响用户 = 小区数 × 单小区在线用户 × 忙闲时系数 × 覆盖补偿系数;受影响用户·小时 = 受影响用户 × 退服时长;业务损失 = 小区话务 × 忙闲时系数 × 覆盖补偿系数 × 退服时长。覆盖补偿:500 米内每有一个邻站下降 10%(室分 5%),最低 40%。
## 测试
测试都在 `tests/_work/` 下建临时数据库,不影响 `data/outage.db`。
```bash
python tests/test_rootcause.py # 根因判定准确率:基线 + 6 组扰动数据(去父告警、时钟偏移、告警缺失、干扰告警、提前量变宽、综合)
python tests/test_kpi.py # 统计周期、稀疏筛选、月均口径、看板与导出一致、影响评估、关联量化
python tests/test_api.py # 登录与权限、导入模板校验、闭环流程与响应时限、站型系数、治理成效、无数据边界
python bench.py # 单批 1 万条导入耗时、全量分析耗时、时长计算准确率
```
根因测试的时间窗只用 1~6 月流水标定,在 7~9 月流水上计分(留出检验)。数据为合成数据,准确率反映的是算法在设定扰动下的表现,论文中应同时说明扰动设置。
+105 -29
View File
@@ -5,10 +5,10 @@
接口分组(与下方代码顺序一致): 接口分组(与下方代码顺序一致):
页面与认证 / /login /logout /healthz 页面与认证 / /login /logout /healthz
看板与统计 /api/meta /api/dashboard /api/stats /api/confusion /api/sites/rank 看板与统计 /api/meta /api/dashboard /api/stats /api/confusion /api/association /api/sites/rank /api/effect
退服流水 /api/outages(查询、新增、修改、删除、批量) /api/analyze 退服流水 /api/outages(查询、新增、修改、删除、批量) /api/analyze
数据导入 /api/import/preview /api/import/confirm /api/template 数据导入 /api/import/preview /api/import/confirm /api/template
站点与场景 /api/sites /api/scenes 站点与场景 /api/sites /api/scenes /api/site_types
导出与系统 /api/export/<报表> /api/logs /api/users 导出与系统 /api/export/<报表> /api/logs /api/users
""" """
import io import io
@@ -144,7 +144,7 @@ def meta():
overview = query("SELECT (SELECT COUNT(*) FROM sites) AS sites, COUNT(*) AS records, " overview = query("SELECT (SELECT COUNT(*) FROM sites) AS sites, COUNT(*) AS records, "
"SUM(status <> '已闭环') AS open FROM outages")[0] # 侧边栏底部的数据概况 "SUM(status <> '已闭环') AS open FROM outages")[0] # 侧边栏底部的数据概况
return jsonify(scenes=core.SCENES, site_types=core.SITE_TYPES, causes=core.CAUSES, months=months, return jsonify(scenes=core.SCENES, site_types=core.SITE_TYPES, causes=core.CAUSES, months=months,
regions=regions, vendors=vendors, status=['待处理', '处理中', '已闭环'], regions=regions, vendors=vendors, status=core.STATUSES,
role=session['role'], user=session['user'], overview=overview) role=session['role'], user=session['user'], overview=overview)
@@ -154,7 +154,8 @@ def dashboard():
started = time.time() started = time.time()
filters = current_filters() filters = current_filters()
df = core.frame(filters) df = core.frame(filters)
stats = core.site_stats(df) minutes, months = core.period(filters)
stats = core.site_stats(df, months)
# 月度趋势:次数与时长(小时) # 月度趋势:次数与时长(小时)
trend = df.groupby('month').agg(n=('id', 'size'), h=('duration_min', 'sum')).reset_index() trend = df.groupby('month').agg(n=('id', 'size'), h=('duration_min', 'sum')).reset_index()
@@ -175,11 +176,12 @@ def dashboard():
scene_sites = core.site_count({**filters, 'scene': scene}) scene_sites = core.site_count({**filters, 'scene': scene})
if scene_sites == 0: if scene_sites == 0:
continue continue
scene_kpi = core.kpis(df[df['scene'] == scene], scene_sites) scene_kpi = core.kpis(df[df['scene'] == scene], scene_sites, minutes, months)
bench.append(dict(scene=scene, **scene_kpi)) bench.append(dict(scene=scene, **scene_kpi))
return jsonify( return jsonify(
kpi=core.kpis(df, core.site_count(filters)), kpi=core.kpis(df, core.site_count(filters), minutes, months),
period=dict(days=round(minutes / 1440, 1), months=months),
elapsed=round(time.time() - started, 3), elapsed=round(time.time() - started, 3),
rank=to_records(stats.head(15), ['site_name', 'scene', 'total_min', 'count']), rank=to_records(stats.head(15), ['site_name', 'scene', 'total_min', 'count']),
trend=trend.to_dict('records'), trend=trend.to_dict('records'),
@@ -209,10 +211,10 @@ def stats():
grouped = df.groupby(dimension).agg( grouped = df.groupby(dimension).agg(
count=('id', 'size'), total_min=('duration_min', 'sum'), avg_min=('duration_min', 'mean'), count=('id', 'size'), total_min=('duration_min', 'sum'), avg_min=('duration_min', 'mean'),
max_min=('duration_min', 'max'), long_count=('is_long', 'sum'), users=('affected_users', 'sum'), max_min=('duration_min', 'max'), long_count=('is_long', 'sum'), users=('affected_users', 'sum'),
loss_gb=('loss_gb', 'sum'), sites=('site_name', 'nunique'), user_h=('user_hours', 'sum'), loss_gb=('loss_gb', 'sum'), sites=('site_name', 'nunique'),
).reset_index() ).reset_index()
grouped['long_ratio'] = (grouped['long_count'] / grouped['count'] * 100).round(2) grouped['long_ratio'] = (grouped['long_count'] / grouped['count'] * 100).round(2)
grouped = grouped.round({'total_min': 1, 'avg_min': 1, 'loss_gb': 1}) grouped = grouped.round({'total_min': 1, 'avg_min': 1, 'loss_gb': 1, 'user_h': 1})
sort = request.args.get('sort', 'total_min') sort = request.args.get('sort', 'total_min')
if sort not in grouped.columns: if sort not in grouped.columns:
@@ -231,14 +233,40 @@ def confusion():
df = df[df['cause_category'].isin(core.CAUSES)] df = df[df['cause_category'].isin(core.CAUSES)]
matrix = pd.crosstab(df['cause_category'], df['root_cause']) matrix = pd.crosstab(df['cause_category'], df['root_cause'])
matrix = matrix.reindex(index=core.CAUSES, columns=core.CAUSES, fill_value=0) matrix = matrix.reindex(index=core.CAUSES, columns=core.CAUSES, fill_value=0)
return jsonify(causes=core.CAUSES, matrix=matrix.values.tolist()) # 按置信度分组的准确率:用来说明“高 / 中 / 低”置信度标签是否可信
by_confidence = []
for level in ['高', '中', '低']:
part = df[df['confidence'] == level]
if len(part):
by_confidence.append(dict(level=level, n=len(part),
acc=round((part['cause_category'] == part['root_cause']).mean() * 100, 2)))
return jsonify(causes=core.CAUSES, matrix=matrix.values.tolist(), by_confidence=by_confidence)
@app.route('/api/association')
@login_required
def association():
"""告警类别 → 退服 的关联强度(支持度、置信度、提升度)与当前使用的时间窗。"""
return jsonify(rows=core.association(current_filters()), windows=core.window_table())
@app.route('/api/effect')
@login_required
def effect():
"""治理成效评估:split=YYYY-MM-DD 为治理分界日期,不传时取评估周期中点所在月的 1 日。"""
try:
result = core.effect(current_filters(), request.args.get('split') or None)
except ValueError as e:
return jsonify(error=str(e)), 400
return jsonify(result)
@app.route('/api/sites/rank') @app.route('/api/sites/rank')
@login_required @login_required
def site_rank(): def site_rank():
"""超长退服站点(kind=long)/ 高频重复站点(kind=repeat)清单,附处置建议。""" """超长退服站点(kind=long)/ 高频重复站点(kind=repeat)清单,附处置建议。"""
stats = core.site_stats(core.frame(current_filters())) filters = current_filters()
stats = core.site_stats(core.frame(filters), core.period(filters)[1])
kind = request.args.get('kind', 'long') kind = request.args.get('kind', 'long')
if kind == 'long': if kind == 'long':
stats = stats[stats['long']] stats = stats[stats['long']]
@@ -254,8 +282,9 @@ def site_rank():
# ================================================================ 退服流水管理 # ================================================================ 退服流水管理
OUTAGE_LIST_COLUMNS = ['id', 'record_id', 'ne_id', 'cell_id', 'site_name', 'scene', 'site_type', 'region', 'vendor', OUTAGE_LIST_COLUMNS = ['id', 'record_id', 'ne_id', 'cell_id', 'site_name', 'scene', 'site_type', 'region', 'vendor',
'start_time', 'end_time', 'duration_min', 'reported_min', 'dur_diff', 'cause_category', 'start_time', 'end_time', 'duration_min', 'reported_min', 'dur_diff', 'cause_category',
'cause_detail', 'cause_code', 'ticket_id', 'root_cause', 'evidence', 'level', 'status', 'cause_detail', 'cause_code', 'ticket_id', 'root_cause', 'evidence', 'confidence', 'level',
'handler', 'remark', 'affected_users', 'loss_gb', 'advice'] 'status', 'handler', 'remark', 'affected_users', 'user_hours', 'loss_gb', 'advice',
'handled_at', 'closed_at', 'response_min', 'close_h', 'limit_h', 'sla']
# 新增 / 管理员修改时写入的字段,顺序与 outage_values() 返回值一致 # 新增 / 管理员修改时写入的字段,顺序与 outage_values() 返回值一致
OUTAGE_EDIT_FIELDS = ['record_id', 'ne_id', 'cell_id', 'start_time', 'end_time', 'duration_min', 'cause_category', OUTAGE_EDIT_FIELDS = ['record_id', 'ne_id', 'cell_id', 'start_time', 'end_time', 'duration_min', 'cause_category',
@@ -265,13 +294,15 @@ OUTAGE_EDIT_FIELDS = ['record_id', 'ne_id', 'cell_id', 'start_time', 'end_time',
@app.route('/api/outages') @app.route('/api/outages')
@login_required @login_required
def outages(): def outages():
"""退服流水分页查询。除通用筛选外,long=1 只看超长,diff=1 只看上报时长差错。""" """退服流水分页查询。除通用筛选外,long=1 只看超长,diff=1 只看上报时长差错,overdue=1 只看超时未闭环。"""
filters = current_filters() filters = current_filters()
df = core.frame(filters) df = core.frame(filters)
if request.args.get('long') == '1': if request.args.get('long') == '1':
df = df[df['is_long']] df = df[df['is_long']]
if request.args.get('diff') == '1': if request.args.get('diff') == '1':
df = df[df['dur_diff']] df = df[df['dur_diff']]
if request.args.get('overdue') == '1':
df = df[(df['status'] != '已闭环') & df['sla'].isin(['超时', '超时未响应'])]
sort = request.args.get('sort', 'start_time') sort = request.args.get('sort', 'start_time')
if sort not in df.columns: if sort not in df.columns:
@@ -296,6 +327,8 @@ def outage_values(data):
raise ValueError('恢复时间早于开始时间') raise ValueError('恢复时间早于开始时间')
if not query('SELECT 1 FROM sites WHERE ne_id=?', (data['ne_id'],)): if not query('SELECT 1 FROM sites WHERE ne_id=?', (data['ne_id'],)):
raise ValueError('基站不在站点库中') raise ValueError('基站不在站点库中')
if (data.get('status') or '待处理') not in core.STATUSES:
raise ValueError('闭环状态不正确')
values = { values = {
'record_id': data.get('record_id', ''), 'record_id': data.get('record_id', ''),
@@ -319,7 +352,13 @@ def outage_values(data):
def outage_add(): def outage_add():
columns = ','.join(OUTAGE_EDIT_FIELDS) columns = ','.join(OUTAGE_EDIT_FIELDS)
placeholders = ','.join('?' * len(OUTAGE_EDIT_FIELDS)) placeholders = ','.join('?' * len(OUTAGE_EDIT_FIELDS))
new_id = execute(f'INSERT INTO outages({columns}) VALUES({placeholders})', outage_values(request.get_json())) data = request.get_json()
try:
values = outage_values(data)
except ValueError as e:
return jsonify(error=str(e)), 400
new_id = execute(f'INSERT INTO outages({columns}) VALUES({placeholders})', values)
core.update_status([new_id], data.get('status') or '待处理') # 按状态补接单 / 闭环时间
core.analyze_all([new_id]) core.analyze_all([new_id])
log(session['user'], '新增退服记录', f'id={new_id}') log(session['user'], '新增退服记录', f'id={new_id}')
return jsonify(ok=True, id=new_id) return jsonify(ok=True, id=new_id)
@@ -330,20 +369,29 @@ def outage_add():
def outage_update(outage_id): def outage_update(outage_id):
""" """
两种修改: 两种修改:
闭环处理:普通用户(或请求中没有 ne_id 时)只更新状态、处理人、处理说明; 闭环处理:普通用户(或请求中没有 ne_id 时)只更新状态、处理人、处理说明,接单 / 闭环时间自动记录;
完整修改:管理员修改全部字段,并重新做根因分析。 完整修改:管理员修改全部字段,并重新做根因分析。
""" """
data = request.get_json() data = request.get_json()
is_admin = session['role'] == 'admin' is_admin = session['role'] == 'admin'
if not query('SELECT 1 FROM outages WHERE id=?', (outage_id,)):
return jsonify(error='记录不存在'), 404
if not is_admin or 'ne_id' not in data: if not is_admin or 'ne_id' not in data:
execute('UPDATE outages SET status=?, handler=?, remark=? WHERE id=?', try:
(data.get('status'), data.get('handler', ''), data.get('remark', ''), outage_id)) core.update_status([outage_id], data.get('status') or '', data.get('handler', ''), data.get('remark', ''))
except ValueError as e:
return jsonify(error=str(e)), 400
log(session['user'], '更新闭环进度', f"id={outage_id} -> {data.get('status')}") log(session['user'], '更新闭环进度', f"id={outage_id} -> {data.get('status')}")
return jsonify(ok=True) return jsonify(ok=True)
try:
values = outage_values(data)
except ValueError as e:
return jsonify(error=str(e)), 400
assignments = ','.join(f'{field}=?' for field in OUTAGE_EDIT_FIELDS) assignments = ','.join(f'{field}=?' for field in OUTAGE_EDIT_FIELDS)
execute(f'UPDATE outages SET {assignments} WHERE id=?', outage_values(data) + (outage_id,)) execute(f'UPDATE outages SET {assignments} WHERE id=?', values + (outage_id,))
core.update_status([outage_id], data.get('status') or '待处理')
core.analyze_all([outage_id]) core.analyze_all([outage_id])
log(session['user'], '修改退服记录', f'id={outage_id}') log(session['user'], '修改退服记录', f'id={outage_id}')
return jsonify(ok=True) return jsonify(ok=True)
@@ -372,7 +420,10 @@ def outage_batch():
if data.get('action') == 'delete': if data.get('action') == 'delete':
execute(f'DELETE FROM outages WHERE id IN ({placeholders})', ids) execute(f'DELETE FROM outages WHERE id IN ({placeholders})', ids)
else: else:
execute(f'UPDATE outages SET status=? WHERE id IN ({placeholders})', [data.get('status', '已闭环')] + ids) try:
core.update_status(ids, data.get('status', '已闭环'))
except ValueError as e:
return jsonify(error=str(e)), 400
log(session['user'], '批量维护', f"{data.get('action')} {len(ids)} 条") log(session['user'], '批量维护', f"{data.get('action')} {len(ids)} 条")
return jsonify(ok=True, n=len(ids)) return jsonify(ok=True, n=len(ids))
@@ -380,12 +431,12 @@ def outage_batch():
@app.route('/api/analyze', methods=['POST']) @app.route('/api/analyze', methods=['POST'])
@admin_required @admin_required
def reanalyze(): def reanalyze():
"""对全部退服流水重新做根因分析(修改阈值或导入告警后使用)。""" """先按历史流水重新标定时间窗,再对全部退服流水重新做根因分析(导入告警或流水后使用)。"""
started = time.time() started = time.time()
count = core.analyze_all() count = core.analyze_all()
accuracy = core.kpis(core.frame())['acc'] accuracy = core.kpis(core.frame())['acc']
log(session['user'], '全量根因分析', f'{count} 条') log(session['user'], '全量根因分析', f'{count} 条')
return jsonify(n=count, elapsed=round(time.time() - started, 3), acc=accuracy) return jsonify(n=count, elapsed=round(time.time() - started, 3), acc=accuracy, windows=core.window_table())
# ================================================================ 数据导入 # ================================================================ 数据导入
@@ -489,7 +540,7 @@ def template():
@login_required @login_required
def sites(): def sites():
return jsonify(rows=query('SELECT * FROM sites ORDER BY scene, district, site_name'), return jsonify(rows=query('SELECT * FROM sites ORDER BY scene, district, site_name'),
scenes=query('SELECT * FROM scenes')) scenes=query('SELECT * FROM scenes'), site_types=query('SELECT * FROM site_types'))
@app.route('/api/sites/<site_id>', methods=['PUT']) @app.route('/api/sites/<site_id>', methods=['PUT'])
@@ -513,6 +564,22 @@ def scene_edit(scene):
return jsonify(ok=True) return jsonify(ok=True)
@app.route('/api/site_types/<site_type>', methods=['PUT'])
@admin_required
def site_type_edit(site_type):
"""站型阈值系数:站点实际阈值 = 场景阈值 × 系数。"""
if site_type not in core.SITE_TYPES:
return jsonify(error='未知站型'), 404
data = request.get_json()
long_factor = float(data['long_factor'])
freq_factor = float(data['freq_factor'])
if not (0.1 <= long_factor <= 10 and 0.1 <= freq_factor <= 10):
return jsonify(error='系数应在 0.1 ~ 10 之间'), 400
execute('INSERT OR REPLACE INTO site_types VALUES(?,?,?)', (site_type, long_factor, freq_factor))
log(session['user'], '修改站型系数', f'{site_type}: 时长×{long_factor} 次数×{freq_factor}')
return jsonify(ok=True)
# ================================================================ 报表导出 # ================================================================ 报表导出
# 导出 Excel 时的中文列名 # 导出 Excel 时的中文列名
EXPORT_COLUMN_NAMES = { EXPORT_COLUMN_NAMES = {
@@ -523,6 +590,9 @@ EXPORT_COLUMN_NAMES = {
'duration_min': '系统计算时长(分)', 'reported_min': '上报时长(分)', 'cause_category': '上报原因大类', 'duration_min': '系统计算时长(分)', 'reported_min': '上报时长(分)', 'cause_category': '上报原因大类',
'cause_detail': '原因细项', 'cause_code': '原因码', 'ticket_id': '工单号', 'root_cause': '系统判定根因', 'cause_detail': '原因细项', 'cause_code': '原因码', 'ticket_id': '工单号', 'root_cause': '系统判定根因',
'evidence': '关联告警', 'level': '断站分级', 'status': '闭环状态', 'advice': '处置建议', 'record_id': '流水号', 'evidence': '关联告警', 'level': '断站分级', 'status': '闭环状态', 'advice': '处置建议', 'record_id': '流水号',
'confidence': '判定置信度', 'user_h': '受影响用户·小时', 'user_hours': '受影响用户·小时',
'handled_at': '接单时间', 'closed_at': '闭环时间', 'response_min': '响应时长(分)', 'limit_h': '响应时限(h)',
'sla': '时限判定',
} }
@@ -546,20 +616,25 @@ def excel_response(sheets, filename):
@login_required @login_required
def export(kind): def export(kind):
"""三类报表:stats 退服统计表、long 超长退服站点清单、advice 根因处置建议清单。""" """三类报表:stats 退服统计表、long 超长退服站点清单、advice 根因处置建议清单。"""
df = core.frame(current_filters()) filters = current_filters()
stats = core.site_stats(df) df = core.frame(filters)
minutes, months = core.period(filters)
stats = core.site_stats(df, months)
stamp = datetime.now().strftime('%Y%m%d%H%M') stamp = datetime.now().strftime('%Y%m%d%H%M')
if kind == 'stats': if kind == 'stats':
k = core.kpis(df) k = core.kpis(df, core.site_count(filters), minutes, months) # 与看板同一口径
kpi_rows = [ kpi_rows = [
['退服次数', k['count']], ['累计退服时长(小时)', k['total_h']], ['平均退服时长(分钟)', k['avg_min']], ['退服次数', k['count']], ['累计退服时长(小时)', k['total_h']], ['平均退服时长(分钟)', k['avg_min']],
['退服率(%)', k['outage_rate']], ['网络可用率(%)', k['availability']], ['超长退服占比(%)', k['long_ratio']], ['退服率(%)', k['outage_rate']], ['网络可用率(%)', k['availability']], ['超长退服占比(%)', k['long_ratio']],
['重复退服率(%)', k['repeat_rate']], ['受影响用户数', k['users']], ['业务损失(GB)', k['loss_gb']], ['重复退服率(%)', k['repeat_rate']], ['受影响用户数', k['users']], ['受影响用户·小时', k['user_h']],
['根因判定准确率(%)', k['acc']], ['闭环率(%)', k['closed']], ['上报时长差错(条)', k['dur_fix']], ['业务损失(GB)', k['loss_gb']], ['统计站点数', k['sites']], ['统计周期(天)', round(minutes / 1440, 1)],
['根因判定准确率(%)', k['acc']], ['闭环率(%)', k['closed']], ['及时响应率(%)', k['response_rate']],
['平均闭环时长(h)', k['avg_close_h']], ['超时未闭环(条)', k['overdue_open']],
['上报时长差错(条)', k['dur_fix']],
] ]
site_columns = ['ne_id', 'site_name', 'region', 'scene', 'site_type', 'count', 'total_min', 'avg_min', site_columns = ['ne_id', 'site_name', 'region', 'scene', 'site_type', 'count', 'total_min', 'avg_min',
'max_min', 'long_count', 'users', 'top_cause'] 'max_min', 'long_count', 'users', 'user_h', 'top_cause']
sheets = { sheets = {
'考核指标': pd.DataFrame(kpi_rows, columns=['指标', '数值']), '考核指标': pd.DataFrame(kpi_rows, columns=['指标', '数值']),
'站点退服统计': stats[site_columns].rename(columns=EXPORT_COLUMN_NAMES), '站点退服统计': stats[site_columns].rename(columns=EXPORT_COLUMN_NAMES),
@@ -589,7 +664,8 @@ def export(kind):
for cause, level in zip(long_outages['root_cause'], long_outages['level'])] for cause, level in zip(long_outages['root_cause'], long_outages['level'])]
long_outages = long_outages.sort_values('duration_min', ascending=False) long_outages = long_outages.sort_values('duration_min', ascending=False)
columns = ['record_id', 'ne_id', 'site_name', 'scene', 'start_time', 'end_time', 'duration_min', columns = ['record_id', 'ne_id', 'site_name', 'scene', 'start_time', 'end_time', 'duration_min',
'cause_category', 'cause_code', 'root_cause', 'evidence', 'level', 'status', 'ticket_id', 'advice'] 'cause_category', 'cause_code', 'root_cause', 'evidence', 'confidence', 'level', 'status',
'handled_at', 'response_min', 'limit_h', 'sla', 'closed_at', 'ticket_id', 'advice']
sheets = {'根因处置建议': long_outages[columns].rename(columns=EXPORT_COLUMN_NAMES)} sheets = {'根因处置建议': long_outages[columns].rename(columns=EXPORT_COLUMN_NAMES)}
title = '根因处置建议清单' title = '根因处置建议清单'
+71
View File
@@ -0,0 +1,71 @@
"""性能与时长计算准确率测试(对应开题报告的预期目标),走 Flask 接口,结果可直接写入论文测试章节。
- 单批 10000 条退服流水导入(预览校验 + 确认入库 + 新记录根因分析)≤ 15 秒
- 单次全量退服时长计算与根因分析 ≤ 8 秒
- 退服时长计算准确率 ≥ 98%(以生成器真值为准),并核对人工上报时长差错的检出
运行:python bench.py 数据库建在 tests/_work 下,不影响 data/outage.db
"""
import io
import os
import sys
import time
sys.path.insert(0, os.path.join(os.path.dirname(os.path.abspath(__file__)), 'tests'))
import harness # noqa: E402 设置数据库路径
from harness import check, core, finish, generate # noqa: E402
if os.path.exists(os.environ['OUTAGE_DB']):
os.remove(os.environ['OUTAGE_DB'])
core.DB_PATH = os.environ['OUTAGE_DB']
started = time.time()
import app as webapp # noqa: E402 导入时 init_db:建库 + 样例数据入库 + 全量分析
print(f'首次建库、样例入库与全量分析: {time.time() - started:.2f} s')
client = webapp.app.test_client()
client.post('/login', json={'username': 'admin', 'password': 'admin123'})
def timed(label, request):
begin = time.time()
response = request()
elapsed = time.time() - begin
print(f'{label}: {elapsed:.2f} s(HTTP {response.status_code})')
return response, elapsed
# 10000 条新流水:同一站点库,取生成序列第 6001~16000 条,避免与样例数据重复
_, alarms_big, outages_big = generate('bench', rows=16000)
batch = outages_big.iloc[6000:16000]
upload = io.BytesIO(batch.to_csv(index=False).encode('utf-8-sig'))
response, t_preview = timed('导入预览 10000 条', lambda: client.post(
'/api/import/preview', data={'file': (upload, 'batch.csv')}, content_type='multipart/form-data'))
preview = response.get_json()
print(f' 合格 {preview["ok"]},问题 {preview["bad"]},上报时长差错 {preview["diff"]}')
response, t_confirm = timed('导入确认(入库 + 新记录根因分析)', lambda: client.post(
'/api/import/confirm', json={'token': preview['token']}))
check('单批 10000 条导入 ≤ 15 s', t_preview + t_confirm <= 15, f'({t_preview + t_confirm:.2f} s)')
upload = io.BytesIO(alarms_big.to_csv(index=False).encode('utf-8-sig'))
response, _ = timed(f'告警预览 {len(alarms_big)} 条', lambda: client.post(
'/api/import/preview', data={'file': (upload, 'alarm.csv')}, content_type='multipart/form-data'))
timed('告警确认(入库 + 全量重新标定与关联)', lambda: client.post(
'/api/import/confirm', json={'token': response.get_json()['token']}))
total = core.query('SELECT COUNT(*) AS n FROM outages')[0]['n']
response, t_analyze = timed(f'全量根因分析 /api/analyze({total} 条,含时间窗标定)',
lambda: client.post('/api/analyze'))
core.data_changed()
response, t_dashboard = timed('看板首次请求(全量时长与指标重算)', lambda: client.get('/api/dashboard'))
check('全量时长计算与根因分析 ≤ 8 s', t_analyze + t_dashboard <= 8, f'({t_analyze + t_dashboard:.2f} s)')
df = core.full_frame()
true_minutes = (core.pd.to_datetime(df['end_time']) - df['start']).dt.total_seconds() / 60
accuracy = ((df['duration_min'] - true_minutes).abs() < 0.05).mean() * 100
check('退服时长计算准确率 ≥ 98%', accuracy >= 98, f'({accuracy:.2f}%)')
injected = (df['reported_min'] - true_minutes).abs() > 1
missed = int((injected & ~df['dur_diff']).sum())
false_alarm = int((~injected & df['dur_diff']).sum())
check('人工上报时长差错全部检出且无误报', missed == 0 and false_alarm == 0,
f'(注入 {int(injected.sum())} 条,标记 {int(df["dur_diff"].sum())} 条)')
finish()
+540 -79
View File
@@ -38,9 +38,10 @@ SCENE_DEFAULT = {
'工业园区': (180, 5, 120, 7), '工业园区': (180, 5, 120, 7),
} }
# 时间窗规则:(告警类别, 判定根因, 告警最多早于退服多少分钟),按优先级从高到低排列。 # 时间窗规则:(告警类别, 判定根因, 默认时间窗:告警最多早于退服多少分钟),按优先级从高到低排列。
# 割接操作最先排除;停电有蓄电池后备、机房高温有过热保护延时,所以窗口放宽; # 割接操作最先排除;停电有蓄电池后备、机房高温有过热保护延时,所以窗口放宽;
# 停电往往同时引起传输告警,因为电源优先级更高,这种情况归为电源故障。 # 停电往往同时引起传输告警,因为电源优先级更高,这种情况归为电源故障。
# 实际使用的窗口由 calibrate_windows() 按历史流水标定(只会放宽,不会小于这里的默认值)。
RULES = [ RULES = [
('割接施工', '割接施工', 30), ('割接施工', '割接施工', 30),
('电源故障', '电源故障', 300), ('电源故障', '电源故障', 300),
@@ -49,6 +50,30 @@ RULES = [
('天馈故障', '天馈故障', 30), ('天馈故障', '天馈故障', 30),
('设备硬件', '设备硬件', 15), ('设备硬件', '设备硬件', 15),
] ]
DEFAULT_WINDOW = {category: window for category, _, window in RULES}
PRIORITY = {category: rank for rank, (category, _, _) in enumerate(RULES)}
MATCH_TOL_MIN = 2 # 第一级:“小区退服”告警与退服开始时间允许相差的分钟数(网管与工单系统时钟不完全同步)
CLEAR_TOL_MIN = 15 # 告警清除与业务恢复相差在此范围内,视为“随业务恢复清除”,作为根因的佐证
SUSPECT_FACTOR = 3 # 第三级:时间窗放宽到几倍仍命中的,判为“疑似”根因(置信度低)
WINDOW_MAX = 720 # 标定后的时间窗上限(分钟)
CALIB_QUANTILE = 0.95 # 标定取历史提前量的分位数
CALIB_MIN_SAMPLES = 30 # 某类样本少于此数时不标定,沿用默认窗口
# 覆盖补偿:退服站点周边 COVER_RADIUS_M 米内每有一个邻站,受影响用户和话务损失按 COVER_STEP 递减(邻站吸收部分话务),
# 最低保留 COVER_FLOOR。室分站主要覆盖室内,室外邻站补偿有限,递减幅度减半。
COVER_RADIUS_M = 500
COVER_STEP = 0.1
COVER_FLOOR = 0.4
# 闭环管控
STATUSES = ['待处理', '处理中', '已闭环']
RESPONSE_HOURS = {'重要': 2, '主要': 4, '次要': 24} # 各断站分级的响应时限(从退服开始算到接单),与 LEVEL_TIP 一致
EFFECT_TOL = 10 # 治理成效:前后变化在 ±10% 以内视为“持平”
# 站型阈值系数:站点实际阈值 = 场景阈值 × 站型系数(超长时长系数, 高频次数系数)。
# 室分多覆盖商场、车站等重点楼宇,超长阈值从严;分布式皮基站覆盖范围小、有宏站兜底,两项阈值放宽。
SITE_TYPE_DEFAULT = {'宏站': (1.0, 1.0), '室分': (0.75, 1.0), '分布式皮基站': (1.5, 1.5)}
ADVICE = { ADVICE = {
'电源故障': ['核查市电及开关电源状态,必要时调度油机发电', '检测蓄电池组容量,老化电池纳入更换计划', '评估加装智能电表与远程动环监控'], '电源故障': ['核查市电及开关电源状态,必要时调度油机发电', '检测蓄电池组容量,老化电池纳入更换计划', '评估加装智能电表与远程动环监控'],
@@ -123,23 +148,46 @@ CREATE TABLE IF NOT EXISTS sites(ne_id TEXT PRIMARY KEY, site_name TEXT, distric
CREATE TABLE IF NOT EXISTS outages(id INTEGER PRIMARY KEY, record_id TEXT, ne_id TEXT, cell_id TEXT, CREATE TABLE IF NOT EXISTS outages(id INTEGER PRIMARY KEY, record_id TEXT, ne_id TEXT, cell_id TEXT,
start_time TEXT, end_time TEXT, duration_min REAL, reported_min REAL, cause_category TEXT, cause_detail TEXT, start_time TEXT, end_time TEXT, duration_min REAL, reported_min REAL, cause_category TEXT, cause_detail TEXT,
cause_code TEXT, ticket_id TEXT, root_cause TEXT, evidence TEXT, status TEXT DEFAULT '待处理', cause_code TEXT, ticket_id TEXT, root_cause TEXT, evidence TEXT, status TEXT DEFAULT '待处理',
handler TEXT, remark TEXT); handler TEXT, remark TEXT, confidence TEXT, handled_at TEXT, closed_at TEXT);
CREATE INDEX IF NOT EXISTS ix_out_ne ON outages(ne_id, start_time); CREATE INDEX IF NOT EXISTS ix_out_ne ON outages(ne_id, start_time);
CREATE TABLE IF NOT EXISTS alarms(alarm_id TEXT PRIMARY KEY, ne_id TEXT, alarm_code TEXT, alarm_name TEXT, CREATE TABLE IF NOT EXISTS alarms(alarm_id TEXT PRIMARY KEY, ne_id TEXT, alarm_code TEXT, alarm_name TEXT,
alarm_level TEXT, category TEXT, start_time TEXT, clear_time TEXT, duration_s INTEGER, cleared TEXT, alarm_level TEXT, category TEXT, start_time TEXT, clear_time TEXT, duration_s INTEGER, cleared TEXT,
parent_alarm_id TEXT); parent_alarm_id TEXT);
CREATE INDEX IF NOT EXISTS ix_alarm_ne ON alarms(ne_id, start_time); CREATE INDEX IF NOT EXISTS ix_alarm_ne ON alarms(ne_id, start_time);
CREATE TABLE IF NOT EXISTS logs(id INTEGER PRIMARY KEY, username TEXT, action TEXT, detail TEXT, ts TEXT); CREATE TABLE IF NOT EXISTS logs(id INTEGER PRIMARY KEY, username TEXT, action TEXT, detail TEXT, ts TEXT);
CREATE TABLE IF NOT EXISTS site_types(site_type TEXT PRIMARY KEY, long_factor REAL, freq_factor REAL);
CREATE TABLE IF NOT EXISTS windows(category TEXT PRIMARY KEY, window_min REAL, default_min REAL, samples INTEGER,
source TEXT, updated_at TEXT);
""" """
def migrate():
"""
旧版数据库升级:outages 补 confidence(根因置信度)、handled_at(接单时间)、closed_at(闭环时间)字段,
站型系数表补默认值。补了 confidence 时返回 True,需要重新做一次全量分析。
"""
needs_analysis = False
with db() as conn:
columns = {row['name'] for row in conn.execute('PRAGMA table_info(outages)')}
for column in ('confidence', 'handled_at', 'closed_at'):
if column not in columns:
conn.execute(f'ALTER TABLE outages ADD COLUMN {column} TEXT')
needs_analysis = needs_analysis or column == 'confidence'
for site_type, (long_factor, freq_factor) in SITE_TYPE_DEFAULT.items():
conn.execute('INSERT OR IGNORE INTO site_types VALUES(?,?,?)', (site_type, long_factor, freq_factor))
return needs_analysis
def init_db(): def init_db():
"""启动时调用:建表;数据库文件是第一次创建时,再写入默认账号、场景阈值,并导入 data/ 下的样例数据。""" """启动时调用:建表;数据库文件是第一次创建时,再写入默认账号、场景阈值,并导入 data/ 下的样例数据。"""
is_first_run = not os.path.exists(DB_PATH) is_first_run = not os.path.exists(DB_PATH)
os.makedirs(os.path.dirname(DB_PATH), exist_ok=True) os.makedirs(os.path.dirname(DB_PATH), exist_ok=True)
with db() as conn: with db() as conn:
conn.executescript(SCHEMA) conn.executescript(SCHEMA)
upgraded = migrate()
if not is_first_run: if not is_first_run:
if upgraded:
analyze_all() # 旧库升级:按新算法重新标定时间窗并补上置信度
return return
now = datetime.now().strftime(TIME_FMT) now = datetime.now().strftime(TIME_FMT)
@@ -217,89 +265,314 @@ def save_sites(df):
return added, updated return added, updated
# ================================================================ 5. 根因关联分析 def update_status(ids, status, handler=None, remark=None):
def analyze(outages):
""" """
两级根因判定。outages 需含 id, ne_id, start_time 三列,返回 [(根因, 证据, id), ...]。 更新闭环状态并自动记录时间:
第一级 衍生关系:同站、同一时刻的“小区退服”告警如果带 parent_alarm_id,直接取父告警的类别 转为“处理中 / 已闭环”且还没有接单时间 → 接单时间 = 当前时间;
(网管已经给出因果链,最可靠); 转为“已闭环”且还没有闭环时间 → 闭环时间 = 当前时间;转为其他状态(重新打开)→ 清空闭环时间。
第二级 时间窗: 否则回溯同站在退服前仍未清除的告警,按 RULES 的优先级取第一个命中的类别; handler / remark 不为 None 时一并更新。
"""
if status not in STATUSES:
raise ValueError(f'闭环状态只能是:{"、".join(STATUSES)}')
ids = [int(i) for i in ids]
if not ids:
return 0
now = datetime.now().strftime(TIME_FMT)
sets = ['status = ?',
"handled_at = CASE WHEN ? <> '待处理' THEN COALESCE(handled_at, ?) ELSE handled_at END",
"closed_at = CASE WHEN ? = '已闭环' THEN COALESCE(closed_at, ?) ELSE NULL END"]
args = [status, status, now, status, now]
if handler is not None:
sets.append('handler = ?')
args.append(handler)
if remark is not None:
sets.append('remark = ?')
args.append(remark)
placeholders = ','.join('?' * len(ids))
with db() as conn:
changed = conn.execute(f'UPDATE outages SET {", ".join(sets)} WHERE id IN ({placeholders})',
args + ids).rowcount
data_changed()
return changed
# ================================================================ 5. 根因关联分析
def read_alarms(ne_ids=None, categories=None):
"""读取告警(可按基站、类别过滤),时间列解析为 datetime,开始时间列改名为 alarm_time。"""
conditions = []
args = []
if ne_ids is not None:
ne_ids = list(ne_ids)
if not ne_ids:
return pd.DataFrame(columns=['alarm_id', 'ne_id', 'alarm_name', 'category', 'alarm_time', 'clear_time',
'parent_alarm_id'])
# 基站很多时(导入全网工参后)不用 IN (?, ?, ...),避免超过 SQLite 的参数个数上限
conditions.append('ne_id IN (SELECT value FROM json_each(?))')
args.append(pd.Series(ne_ids).to_json(orient='values', force_ascii=False))
if categories is not None:
conditions.append('category IN (SELECT value FROM json_each(?))')
args.append(pd.Series(list(categories)).to_json(orient='values', force_ascii=False))
sql = 'SELECT alarm_id, ne_id, alarm_name, category, start_time, clear_time, parent_alarm_id FROM alarms'
if conditions:
sql += ' WHERE ' + ' AND '.join(conditions)
with db() as conn:
alarms = pd.read_sql(sql, conn, params=args)
alarms['start_time'] = pd.to_datetime(alarms['start_time'], errors='coerce', format='mixed')
alarms['clear_time'] = pd.to_datetime(alarms['clear_time'], errors='coerce', format='mixed')
alarms = alarms.dropna(subset=['start_time'])
return alarms.rename(columns={'start_time': 'alarm_time'})
def still_active(alarm_clear, outage_start):
"""告警在退服开始时仍未清除(允许 1 分钟误差)才可能是退服的原因。"""
return alarm_clear.isna() | (alarm_clear >= outage_start - pd.Timedelta(minutes=1))
def calibrate_windows(until=None):
"""
用历史流水标定各类根因告警的时间窗(告警最多早于退服多少分钟),结果写入 windows 表并返回 {类别: 分钟}。
样本:已闭环、上报原因属于规则类别的退服(until 指定时只用该日期之前的,便于留出检验);
在同站、同类别、退服时仍未清除、提前量不超过 WINDOW_MAX 的告警中,取离退服最近的一条,记其提前量;
窗口:提前量的 95% 分位数,不小于默认窗口、不大于 WINDOW_MAX;样本少于 30 条时沿用默认窗口。
这样时间窗随本地网络的实际情况(蓄电池后备时长、动环告警延时等)自动调整,而不是写死在代码里。
"""
categories = list(DEFAULT_WINDOW)
sql = ("SELECT id, ne_id, start_time, cause_category FROM outages WHERE status = '已闭环' "
"AND cause_category IN (SELECT value FROM json_each(?))")
args = [pd.Series(categories).to_json(orient='values', force_ascii=False)]
if until:
sql += ' AND start_time < ?'
args.append(until)
with db() as conn:
outages = pd.read_sql(sql, conn, params=args)
outages['start_time'] = pd.to_datetime(outages['start_time'])
alarms = read_alarms(outages['ne_id'].unique(), categories)
pairs = outages.merge(alarms, left_on=['ne_id', 'cause_category'], right_on=['ne_id', 'category'])
lead = (pairs['start_time'] - pairs['alarm_time']).dt.total_seconds() / 60
usable = still_active(pairs['clear_time'], pairs['start_time']) & (lead >= -5) & (lead <= WINDOW_MAX)
pairs = pairs[usable].assign(lead=lead[usable].clip(lower=0))
nearest = pairs.sort_values('lead').drop_duplicates('id')
now = datetime.now().strftime(TIME_FMT)
result = {}
rows = []
for category in categories:
leads = nearest.loc[nearest['cause_category'] == category, 'lead']
default = DEFAULT_WINDOW[category]
if len(leads) >= CALIB_MIN_SAMPLES:
window = float(min(max(leads.quantile(CALIB_QUANTILE), default), WINDOW_MAX))
source = '历史标定'
else:
window = float(default)
source = '默认(样本不足)'
result[category] = round(window, 1)
rows.append((category, round(window, 1), default, int(len(leads)), source, now))
with db() as conn:
conn.execute('DELETE FROM windows')
conn.executemany('INSERT INTO windows VALUES(?,?,?,?,?,?)', rows)
return result
def load_windows():
"""当前使用的时间窗 {类别: 分钟};未标定的类别用默认窗口。"""
windows = dict(DEFAULT_WINDOW)
for row in query('SELECT category, window_min FROM windows'):
if row['category'] in windows and row['window_min']:
windows[row['category']] = row['window_min']
return windows
def window_table():
"""时间窗明细(页面展示用)。"""
saved = {row['category']: row for row in query('SELECT * FROM windows')}
rows = []
for category, _, default in RULES:
row = saved.get(category) or {'window_min': default, 'samples': 0, 'source': '默认', 'updated_at': ''}
rows.append(dict(category=category, priority=PRIORITY[category] + 1, default_min=default,
window_min=row['window_min'], samples=row['samples'], source=row['source'],
updated_at=row['updated_at']))
return rows
def analyze(outages, windows=None):
"""
三级根因判定。outages 需含 id, ne_id, start_time, end_time 四列,返回 [(根因, 证据, 置信度, id), ...]。
预处理 衍生告警归并:告警带 parent_alarm_id 且父告警属于根因类别时,按父告警的类别和名称计
(如停电引起的传输告警归为电源故障);
第一级 衍生关系(置信度高):同站“小区退服”告警与退服开始相差不超过 MATCH_TOL_MIN 分钟、且带父告警,
直接取父告警类别;
第二级 时间窗(置信度中):同站在退服时仍未清除、提前量在该类时间窗内的告警;
第三级 疑似(置信度低):时间窗放宽到 SUSPECT_FACTOR 倍才命中的告警,提示人工复核;
多条命中时依次比较:级别 → 是否随业务恢复清除(清除与恢复相差 ≤ CLEAR_TOL_MIN 分钟)→ 规则优先级 → 时间早;
都没有命中判为“其他”。 都没有命中判为“其他”。
""" """
if outages.empty: if outages.empty:
return [] return []
windows = windows or load_windows()
# 只取涉及到的基站的告警,避免全表读取 outages = outages[['id', 'ne_id', 'start_time', 'end_time']].copy()
ne_ids = tuple(outages['ne_id'].dropna().unique())
placeholders = ','.join('?' * len(ne_ids))
with db() as conn:
alarms = pd.read_sql(f'SELECT alarm_id, ne_id, alarm_name, category, start_time, clear_time, parent_alarm_id '
f'FROM alarms WHERE ne_id IN ({placeholders})', conn, params=ne_ids,
parse_dates=['start_time', 'clear_time'])
alarms = alarms.rename(columns={'start_time': 'alarm_time'})
outages = outages[['id', 'ne_id', 'start_time']].copy()
outages['start_time'] = pd.to_datetime(outages['start_time']) outages['start_time'] = pd.to_datetime(outages['start_time'])
outages['end_time'] = pd.to_datetime(outages['end_time'])
alarms = read_alarms(outages['ne_id'].dropna().unique())
results = {} # 退服 id -> (根因, 证据) # ---- 预处理:衍生告警按父告警计
alarms['root_cat'] = alarms['category']
alarms['root_name'] = alarms['alarm_name']
parents = alarms.drop_duplicates('alarm_id').set_index('alarm_id')
parent_id = alarms['parent_alarm_id'].fillna('')
parent_cat = parent_id.map(parents['category'])
linked = parent_cat.isin(DEFAULT_WINDOW)
alarms.loc[linked, 'root_cat'] = parent_cat[linked]
alarms.loc[linked, 'root_name'] = parent_id[linked].map(parents['alarm_name'])
# 证据里显示的告警名:衍生告警写成“传输告警(衍生自 市电停电)”
alarms['show_name'] = alarms['alarm_name']
alarms.loc[linked, 'show_name'] = alarms.loc[linked, 'alarm_name'] + '(衍生自 ' + alarms.loc[linked, 'root_name'] + ')'
# ---- 第一级:衍生关系。找到带父告警的“小区退服”告警,再查出父告警的类别 results = {} # 退服 id -> (根因, 证据, 置信度)
child_alarms = alarms[(alarms['category'] == '小区退服') & alarms['parent_alarm_id'].fillna('').ne('')]
child_alarms = child_alarms[['ne_id', 'alarm_time', 'parent_alarm_id']].rename(columns={'alarm_time': 'start_time'}) # ---- 第一级:衍生关系
parent_alarms = alarms.drop(columns=['ne_id', 'parent_alarm_id']) child = alarms[(alarms['category'] == '小区退服') & linked][['ne_id', 'alarm_time', 'root_cat', 'root_name']]
child_with_parent = child_alarms.merge(parent_alarms, left_on='parent_alarm_id', right_on='alarm_id') matched = outages.merge(child, on='ne_id')
matched = outages.merge(child_with_parent, on=['ne_id', 'start_time']).drop_duplicates('id') gap = (matched['alarm_time'] - matched['start_time']).abs()
close_enough = gap <= pd.Timedelta(minutes=MATCH_TOL_MIN)
# 先筛选再赋值且只赋筛选后的部分:给空 DataFrame 赋一个整列 Series 时,pandas 会沿用 Series 的索引凭空生成空行
matched = matched[close_enough].assign(gap=gap[close_enough])
matched = matched.sort_values(['id', 'gap']).drop_duplicates('id')
for row in matched.itertuples(): for row in matched.itertuples():
if row.category in CAUSES: results[row.id] = (row.root_cat, f'衍生关系:{row.root_name} → 小区退服@{row.alarm_time:%m-%d %H:%M}', '高')
evidence = f'衍生关系:{row.alarm_name}@{row.alarm_time:%m-%d %H:%M}'
results[row.id] = (row.category, evidence)
# ---- 第二级:时间窗。把剩余退服与同站的根因类告警两两配对,逐条规则打标 # ---- 第二、三级:时间窗与疑似
remaining = outages[~outages['id'].isin(results)] remaining = outages[~outages['id'].isin(results)]
pairs = remaining.merge(alarms[alarms['category'] != '小区退服'], on='ne_id') candidates = alarms[(alarms['category'] != '小区退服') & alarms['root_cat'].isin(DEFAULT_WINDOW)]
minutes_before = (pairs['start_time'] - pairs['alarm_time']).dt.total_seconds() / 60 pairs = remaining.merge(candidates, on='ne_id')
# 告警在退服时仍未清除(允许 1 分钟误差)才可能是原因 lead = (pairs['start_time'] - pairs['alarm_time']).dt.total_seconds() / 60
still_active = pairs['clear_time'].isna() | (pairs['clear_time'] >= pairs['start_time'] - pd.Timedelta(minutes=1)) window = pairs['root_cat'].map(windows).astype(float)
active = still_active(pairs['clear_time'], pairs['start_time']) & (lead >= -5)
pairs['tier'] = np.select([active & (lead <= window), active & (lead <= window * SUSPECT_FACTOR)], [0, 1], 9)
clear_gap = (pairs['clear_time'] - pairs['end_time']).abs()
pairs['sync'] = clear_gap <= pd.Timedelta(minutes=CLEAR_TOL_MIN)
pairs['unsync'] = ~pairs['sync']
pairs['prio'] = pairs['root_cat'].map(PRIORITY)
pairs['rank'] = 99 # 命中规则的序号,越小优先级越高;99 表示没有命中 best = pairs[pairs['tier'] < 9].sort_values(['id', 'tier', 'unsync', 'prio', 'alarm_time']).drop_duplicates('id')
for rank, (category, _, window) in enumerate(RULES):
hit = ((pairs['category'] == category) & (minutes_before <= window) & (minutes_before >= -5)
& still_active & (pairs['rank'] > rank))
pairs.loc[hit, 'rank'] = rank
# 每条退服取优先级最高、时间最早的一条告警作为根因
best = pairs[pairs['rank'] < 99].sort_values(['id', 'rank', 'alarm_time']).drop_duplicates('id')
for row in best.itertuples(): for row in best.itertuples():
evidence = f'时间窗:{row.alarm_name}@{row.alarm_time:%m-%d %H:%M}' sync_note = ',随业务恢复清除' if row.sync else ''
results[row.id] = (RULES[row.rank][1], evidence) if row.tier == 0:
results[row.id] = (row.root_cat, f'时间窗:{row.show_name}@{row.alarm_time:%m-%d %H:%M}{sync_note}', '中')
else:
results[row.id] = (row.root_cat, f'疑似(超出时间窗):{row.show_name}@{row.alarm_time:%m-%d %H:%M}'
f'{sync_note},建议人工复核', '低')
output = [] output = []
for outage_id in outages['id']: for outage_id in outages['id']:
cause, evidence = results.get(outage_id, ('其他', '时间窗内无关联告警')) cause, evidence, confidence = results.get(outage_id, ('其他', '时间窗内无关联告警', '低'))
output.append((cause, evidence, int(outage_id))) output.append((cause, evidence, confidence, int(outage_id)))
return output return output
def analyze_all(ids=None): def analyze_all(ids=None, recalibrate=True):
"""对指定 id(不传则全部)的退服流水重新做根因分析并写回数据库,返回分析条数。""" """
sql = 'SELECT id, ne_id, start_time FROM outages' 对指定 id 的退服流水(ids=None 表示全部)重新做根因分析并写回数据库,返回分析条数。
if ids: 全量分析前先按历史流水重新标定时间窗(recalibrate=False 时沿用已有窗口)。
sql += ' WHERE id IN (' + ','.join(str(i) for i in ids) + ')' """
if ids is not None and len(ids) == 0:
return 0
if ids is None and recalibrate:
calibrate_windows()
sql = 'SELECT id, ne_id, start_time, end_time FROM outages'
if ids is not None:
sql += ' WHERE id IN (' + ','.join(str(int(i)) for i in ids) + ')'
with db() as conn: with db() as conn:
outages = pd.read_sql(sql, conn) outages = pd.read_sql(sql, conn)
results = analyze(outages) results = analyze(outages)
with db() as conn: with db() as conn:
conn.executemany('UPDATE outages SET root_cause=?, evidence=? WHERE id=?', results) conn.executemany('UPDATE outages SET root_cause=?, evidence=?, confidence=? WHERE id=?', results)
data_changed() data_changed()
return len(results) return len(results)
def association(filters=None):
"""
告警类别与退服的关联强度,规则为“同站出现 X 类告警 → 其后 W 分钟内发生退服”(W 为该类时间窗):
告警数 n_A、命中数 n_AB(其后 W 分钟内同站发生退服的告警数)
支持度 support = 开始前 W 分钟内同站有 X 类告警的退服数 ÷ 退服总数,即 P(X 类告警 | 退服)
置信度 confidence = n_AB ÷ n_A,即 P(退服 | X 类告警)
提升度 lift = 置信度 ÷ P0,P0 = 退服数 × W ÷(站点数 × 统计时长),即随机一个 W 分钟窗口内遇到退服的概率
提升度 > 1 说明该类告警出现后退服概率高于随机水平,数值越大关联越强。
"""
filters = filters or {}
df = frame(filters)
lo, hi = period_bounds(filters)
minutes = max((hi - lo).total_seconds() / 60, 1440)
n_sites = max(site_count(filters), 1)
windows = load_windows()
site_where, site_args = site_conditions(filters)
ne_ids = [row['ne_id'] for row in query('SELECT ne_id FROM sites' + site_where, site_args)]
alarms = read_alarms(ne_ids, list(DEFAULT_WINDOW))
alarms = alarms[(alarms['alarm_time'] >= lo) & (alarms['alarm_time'] < hi)]
outages = df[['id', 'ne_id', 'start']].dropna().sort_values('start')
rows = []
for category, _, _ in RULES:
window = windows[category]
tolerance = pd.Timedelta(minutes=window)
cat_alarms = alarms.loc[alarms['category'] == category, ['ne_id', 'alarm_time']].sort_values('alarm_time')
hits = 0
if len(cat_alarms) and len(outages):
forward = pd.merge_asof(cat_alarms, outages, left_on='alarm_time', right_on='start', by='ne_id',
direction='forward', tolerance=tolerance)
hits = int(forward['start'].notna().sum())
covered = 0
if len(cat_alarms) and len(outages):
backward = pd.merge_asof(outages, cat_alarms, left_on='start', right_on='alarm_time', by='ne_id',
direction='backward', tolerance=tolerance)
covered = int(backward['alarm_time'].notna().sum())
p0 = min(len(outages) * window / (n_sites * minutes), 1) if len(outages) else 0
confidence = hits / len(cat_alarms) if len(cat_alarms) else 0
rows.append(dict(
category=category, window_min=window, alarms=len(cat_alarms), hits=hits,
support=round(covered / len(outages) * 100, 2) if len(outages) else 0,
confidence=round(confidence * 100, 2),
lift=round(confidence / p0, 2) if p0 else 0,
))
return rows
# ================================================================ 6. 指标核算 # ================================================================ 6. 指标核算
def site_cover():
"""
各站覆盖补偿系数 {ne_id: 系数}:COVER_RADIUS_M 米内的邻站数 n,系数 = max(COVER_FLOOR, 1 − COVER_STEP × n),
室分站递减幅度减半。距离按经纬度近似换算(城市范围内误差可忽略);没有经纬度的站系数为 1。
"""
sites = pd.DataFrame(query('SELECT ne_id, site_type, lng, lat FROM sites'))
if sites.empty:
return {}
sites = sites.dropna(subset=['lng', 'lat'])
if sites.empty:
return {}
lat0 = np.radians(sites['lat'].astype(float).mean())
x = sites['lng'].to_numpy(float) * 111320 * np.cos(lat0)
y = sites['lat'].to_numpy(float) * 110540
neighbors = np.zeros(len(sites), dtype=int)
for begin in range(0, len(sites), 1000): # 分块计算,站点很多时不会一次占用大量内存
dx = x[begin:begin + 1000, None] - x[None, :]
dy = y[begin:begin + 1000, None] - y[None, :]
neighbors[begin:begin + 1000] = ((dx * dx + dy * dy) <= COVER_RADIUS_M ** 2).sum(axis=1) - 1
step = np.where(sites['site_type'].eq('室分'), COVER_STEP / 2, COVER_STEP)
factor = np.clip(1 - step * neighbors, COVER_FLOOR, 1)
return dict(zip(sites['ne_id'], factor.round(3)))
def full_frame(): def full_frame():
""" """
全部退服流水 + 站点信息 + 场景阈值,并计算派生字段: 全部退服流水 + 站点信息 + 场景阈值 × 站型系数,并计算派生字段:
is_long 是否超长、dur_diff 上报时长差错、affected_users 受影响用户、loss_gb 业务损失、level 断站分级。 is_long 是否超长、dur_diff 上报时长差错、cover_factor 覆盖补偿系数、affected_users 受影响用户、
user_hours 受影响用户·小时、loss_gb 业务损失、level 断站分级、
response_min 响应时长、close_h 闭环时长、limit_h 响应时限、sla 响应时限判定。
结果缓存到数据变化为止(见 data_changed)。 结果缓存到数据变化为止(见 data_changed)。
""" """
global _frame_cache global _frame_cache
@@ -308,22 +581,34 @@ def full_frame():
with db() as conn: with db() as conn:
df = pd.read_sql('SELECT o.*, s.site_name, s.district AS region, s.scene, s.site_type, s.vendor, s.cells, ' df = pd.read_sql('SELECT o.*, s.site_name, s.district AS region, s.scene, s.site_type, s.vendor, s.cells, '
'c.long_min, c.freq_count, c.users_per_cell, c.gb_per_cell_h FROM outages o ' 'c.long_min AS scene_long_min, c.freq_count AS scene_freq_count, c.users_per_cell, '
'LEFT JOIN sites s ON o.ne_id = s.ne_id LEFT JOIN scenes c ON s.scene = c.scene', conn) 'c.gb_per_cell_h, t.long_factor, t.freq_factor FROM outages o '
'LEFT JOIN sites s ON o.ne_id = s.ne_id LEFT JOIN scenes c ON s.scene = c.scene '
'LEFT JOIN site_types t ON s.site_type = t.site_type', conn)
df['site_name'] = df['site_name'].fillna(df['ne_id']) # 站点库里没有的站,用站号代替站名 df['site_name'] = df['site_name'].fillna(df['ne_id']) # 站点库里没有的站,用站号代替站名
df['start'] = pd.to_datetime(df['start_time']) df['start'] = pd.to_datetime(df['start_time'])
df['month'] = df['start_time'].str[:7] df['month'] = df['start_time'].str[:7]
long_min = df['long_min'].fillna(120) # 站点实际阈值 = 场景阈值 × 站型系数
df['long_min'] = (df['scene_long_min'].fillna(120) * df['long_factor'].fillna(1)).round(1)
df['freq_count'] = (df['scene_freq_count'].fillna(4) * df['freq_factor'].fillna(1)).round(1)
long_min = df['long_min']
df['is_long'] = df['duration_min'] >= long_min df['is_long'] = df['duration_min'] >= long_min
df['dur_diff'] = (df['reported_min'] - df['duration_min']).abs() > 1 # 人工上报时长与系统计算相差超过 1 分钟 df['dur_diff'] = (df['reported_min'] - df['duration_min']).abs() > 1 # 人工上报时长与系统计算相差超过 1 分钟
# 影响评估:受影响用户 = 小区数 × 单小区在线用户 × 忙闲时系数;业务损失 = 小区数 × 单小区话务 × 系数 × 时长 # 影响评估:
# 受影响用户 = 小区数 × 单小区在线用户 × 忙闲时系数 × 覆盖补偿系数(同一时刻掉线的用户数)
# 受影响用户·小时 = 受影响用户 × 退服时长(小时),反映影响的持续程度
# 业务损失 = 小区数 × 单小区每小时话务 × 忙闲时系数 × 覆盖补偿系数 × 退服时长(小时)
busy_factor = np.where(df['start'].dt.hour.between(9, 22), 1.0, 0.35) # 9-22 点为忙时 busy_factor = np.where(df['start'].dt.hour.between(9, 22), 1.0, 0.35) # 9-22 点为忙时
cells = df['cells'].fillna(3).astype(float) cells = df['cells'].fillna(3).astype(float)
df['affected_users'] = (cells * df['users_per_cell'].fillna(200) * busy_factor).round().astype(int) df['cover_factor'] = df['ne_id'].map(site_cover()).fillna(1.0).astype(float)
df['loss_gb'] = (cells * df['gb_per_cell_h'].fillna(10) * busy_factor * df['duration_min'] / 60).round(2) hours = df['duration_min'] / 60
users = cells * df['users_per_cell'].fillna(200) * busy_factor * df['cover_factor']
df['affected_users'] = users.round().astype(int)
df['user_hours'] = (users * hours).round(1)
df['loss_gb'] = (cells * df['gb_per_cell_h'].fillna(10) * busy_factor * df['cover_factor'] * hours).round(2)
# 断站分级(按整列一次计算,比逐行判断快几十倍): # 断站分级(按整列一次计算,比逐行判断快几十倍):
# 重要:超长,且是关键场景或时长超过阈值 2 倍 # 重要:超长,且是关键场景或时长超过阈值 2 倍
@@ -334,6 +619,25 @@ def full_frame():
is_major = df['is_long'] | (key_scene & (df['duration_min'] >= 30)) is_major = df['is_long'] | (key_scene & (df['duration_min'] >= 30))
df['level'] = np.select([is_important, is_major], ['重要', '主要'], default='次要') df['level'] = np.select([is_important, is_major], ['重要', '主要'], default='次要')
# 闭环时限(从退服开始算):
# 及时 已接单,响应时长 ≤ 时限
# 超时 已接单,响应时长 > 时限
# 超时未响应 未接单且已超过时限
# 未到期 未接单,还在时限内
# 无记录 已闭环但没有接单时间(历史数据),不参与及时率统计
handled = pd.to_datetime(df['handled_at'], errors='coerce')
closed = pd.to_datetime(df['closed_at'], errors='coerce')
df['response_min'] = ((handled - df['start']).dt.total_seconds() / 60).round(1)
df['close_h'] = ((closed - df['start']).dt.total_seconds() / 3600).round(2)
df['limit_h'] = df['level'].map(RESPONSE_HOURS).astype(float)
waited_min = (pd.Timestamp.now() - df['start']).dt.total_seconds() / 60
df['sla'] = np.select(
[handled.notna() & (df['response_min'] <= df['limit_h'] * 60),
handled.notna(),
(df['status'] != '已闭环') & (waited_min > df['limit_h'] * 60),
df['status'] != '已闭环'],
['及时', '超时', '超时未响应', '未到期'], default='无记录')
_frame_cache = df _frame_cache = df
return df return df
@@ -362,8 +666,49 @@ def frame(filters=None):
return df.copy() return df.copy()
def site_count(filters=None): def period_bounds(filters=None):
"""满足站点类筛选条件(场景、片区、站型、厂家)的站点数,用作退服率的分母。""" """
统计周期 [起, 止)。有月份 / 日期范围筛选时取筛选范围(同时有则取交集),
否则取全部退服流水首日 0 点到末日 24 点。
"""
filters = filters or {}
df = full_frame()
if df.empty:
today = pd.Timestamp.now().normalize()
lo, hi = today, today + pd.Timedelta(days=1)
else:
lo = df['start'].min().normalize()
hi = df['start'].max().normalize() + pd.Timedelta(days=1)
lowers = []
uppers = []
if filters.get('month'):
month_start = pd.Timestamp(filters['month'] + '-01')
lowers.append(month_start)
uppers.append(month_start + pd.offsets.MonthBegin(1))
if filters.get('date_from'):
lowers.append(pd.Timestamp(filters['date_from']))
if filters.get('date_to'):
uppers.append(pd.Timestamp(filters['date_to']) + pd.Timedelta(days=1))
if lowers:
lo = max(lowers)
if uppers:
hi = min(uppers)
if hi <= lo:
hi = lo + pd.Timedelta(days=1)
return lo, hi
def period(filters=None):
"""统计周期的 (分钟数, 自然月数):分钟数是退服率的时间分母,月数是“月均”的分母。"""
lo, hi = period_bounds(filters)
minutes = (hi - lo).total_seconds() / 60
months = len(pd.period_range(lo, hi - pd.Timedelta(seconds=1), freq='M'))
return minutes, max(months, 1)
def site_conditions(filters=None):
"""站点类筛选(场景、片区、站型、厂家、站名/站号关键字)对应的 WHERE 子句与参数。"""
filters = filters or {} filters = filters or {}
column_of = {'scene': 'scene', 'region': 'district', 'site_type': 'site_type', 'vendor': 'vendor'} column_of = {'scene': 'scene', 'region': 'district', 'site_type': 'site_type', 'vendor': 'vendor'}
conditions = [] conditions = []
@@ -372,30 +717,45 @@ def site_count(filters=None):
if filters.get(key): if filters.get(key):
conditions.append(f'{column} = ?') conditions.append(f'{column} = ?')
args.append(filters[key]) args.append(filters[key])
if filters.get('site'):
sql = 'SELECT COUNT(*) AS n FROM sites' conditions.append('(instr(site_name, ?) > 0 OR instr(ne_id, ?) > 0)') # 与 frame() 一样区分大小写
if conditions: args += [filters['site'], filters['site']]
sql += ' WHERE ' + ' AND '.join(conditions) if not conditions:
return query(sql, args)[0]['n'] return '', []
return ' WHERE ' + ' AND '.join(conditions), args
def kpis(df, n_sites=None): def site_count(filters=None):
"""满足站点类筛选条件的站点数,用作退服率、重复退服率的分母。"""
where, args = site_conditions(filters)
return query('SELECT COUNT(*) AS n FROM sites' + where, args)[0]['n']
def kpis(df, n_sites=None, minutes=None, months=None):
""" """
运维考核指标。 运维考核指标。n_sites / minutes / months 为站点数、统计时长(分钟)、统计月数,不传时按全量计算。
退服率 = 总退服时长 ÷(站点数 × 统计时长);网络可用率 = 1 − 退服率 退服率 = 总退服时长 ÷(站点数 × 统计时长);网络可用率 = 1 − 退服率
重复退服率 = 高频重复站点数 ÷ 站点数
及时响应率 = 响应及时数 ÷(及时 + 超时 + 超时未响应);平均闭环时长 = 已闭环且有闭环时间的记录从退服开始到闭环
超时未闭环 = 未闭环且已超过响应时限(含已接单但超时的)
根因判定准确率 = 上报原因与系统判定一致的比例(只统计上报原因已填写的记录) 根因判定准确率 = 上报原因与系统判定一致的比例(只统计上报原因已填写的记录)
""" """
if not n_sites: if n_sites is None:
n_sites = site_count() n_sites = site_count()
if minutes is None or months is None:
default_minutes, default_months = period()
minutes = minutes or default_minutes
months = months or default_months
if df.empty: if df.empty:
return dict(count=0, total_h=0, avg_min=0, outage_rate=0, availability=100, long_ratio=0, repeat_rate=0, return dict(count=0, total_h=0, avg_min=0, outage_rate=0, availability=100, long_ratio=0, repeat_rate=0,
users=0, loss_gb=0, sites=n_sites, acc=0, dur_fix=0, closed=0) users=0, user_h=0, loss_gb=0, sites=n_sites, acc=0, dur_fix=0, closed=0, response_rate=None,
avg_close_h=None, overdue_open=0)
total_minutes = df['duration_min'].sum() total_minutes = df['duration_min'].sum()
period_minutes = max((df['start'].max() - df['start'].min()).total_seconds() / 60, 1) denominator = max(n_sites, 1) * max(minutes, 1)
outage_rate = total_minutes / (n_sites * period_minutes) * 100 outage_rate = total_minutes / denominator * 100
stats = site_stats(df) stats = site_stats(df, months)
known = df[df['cause_category'].isin(CAUSES) & df['root_cause'].notna()] known = df[df['cause_category'].isin(CAUSES) & df['root_cause'].notna()]
if len(known): if len(known):
@@ -403,6 +763,13 @@ def kpis(df, n_sites=None):
else: else:
accuracy = 0 accuracy = 0
judged = df['sla'].isin(['及时', '超时', '超时未响应'])
# 没有可判定的记录(例如旧数据没有接单时间)时返回 None,页面显示“—”,避免误读成 0%
response_rate = round((df['sla'] == '及时').sum() / judged.sum() * 100, 2) if judged.any() else None
close_hours = df.loc[df['status'] == '已闭环', 'close_h'].dropna()
avg_close_h = round(float(close_hours.mean()), 1) if len(close_hours) else None
overdue_open = int(((df['status'] != '已闭环') & df['sla'].isin(['超时', '超时未响应'])).sum())
return dict( return dict(
count=len(df), count=len(df),
total_h=round(total_minutes / 60, 1), total_h=round(total_minutes / 60, 1),
@@ -410,29 +777,38 @@ def kpis(df, n_sites=None):
outage_rate=round(outage_rate, 4), outage_rate=round(outage_rate, 4),
availability=round(100 - outage_rate, 4), availability=round(100 - outage_rate, 4),
long_ratio=round(df['is_long'].mean() * 100, 2), long_ratio=round(df['is_long'].mean() * 100, 2),
repeat_rate=round(stats['repeat'].mean() * 100, 2), repeat_rate=round(stats['repeat'].sum() / max(n_sites, 1) * 100, 2),
users=int(df['affected_users'].sum()), users=int(df['affected_users'].sum()),
user_h=int(round(float(df['user_hours'].sum()))),
loss_gb=round(df['loss_gb'].sum(), 1), loss_gb=round(df['loss_gb'].sum(), 1),
sites=n_sites, sites=n_sites,
acc=accuracy, acc=accuracy,
dur_fix=int(df['dur_diff'].sum()), dur_fix=int(df['dur_diff'].sum()),
closed=round((df['status'] == '已闭环').mean() * 100, 2), closed=round((df['status'] == '已闭环').mean() * 100, 2),
response_rate=response_rate,
avg_close_h=avg_close_h,
overdue_open=overdue_open,
) )
def site_stats(df): def site_stats(df, months=None):
""" """
站点维度统计,按累计退服时长从大到小排序。 站点维度统计,按累计退服时长从大到小排序。months 为统计周期的自然月数,不传时按全量周期。
超长退服站点 long:月均超长退服 ≥ 1 次 超长退服站点 long:月均超长退服 ≥ 1 次
高频重复站点 repeat:月均退服次数 ≥ 场景次数阈值 高频重复站点 repeat:月均退服次数 ≥ 场景次数阈值
""" """
if df.empty: if df.empty:
return pd.DataFrame(columns=['site_name', 'count', 'total_min', 'long', 'repeat']) return pd.DataFrame(columns=['ne_id', 'site_name', 'region', 'scene', 'site_type', 'count', 'total_min',
'avg_min', 'max_min', 'long_count', 'users', 'user_h', 'long_min', 'freq_count',
'top_cause', 'long', 'repeat']).astype({'long': bool, 'repeat': bool})
# long / repeat 必须是布尔列:空的 object 列用作 stats[stats['long']] 时会被当成“选列”,结果丢掉所有列
if months is None:
months = period()[1]
stats = df.groupby(['ne_id', 'site_name', 'region', 'scene', 'site_type'], dropna=False).agg( stats = df.groupby(['ne_id', 'site_name', 'region', 'scene', 'site_type'], dropna=False).agg(
count=('id', 'size'), total_min=('duration_min', 'sum'), avg_min=('duration_min', 'mean'), count=('id', 'size'), total_min=('duration_min', 'sum'), avg_min=('duration_min', 'mean'),
max_min=('duration_min', 'max'), long_count=('is_long', 'sum'), users=('affected_users', 'sum'), max_min=('duration_min', 'max'), long_count=('is_long', 'sum'), users=('affected_users', 'sum'),
long_min=('long_min', 'first'), freq_count=('freq_count', 'first'), user_h=('user_hours', 'sum'), long_min=('long_min', 'first'), freq_count=('freq_count', 'first'),
).reset_index() ).reset_index()
# 主要根因:每站出现次数最多的根因;次数相同时取名称排序靠前的(与 pandas 的 mode 结果一致) # 主要根因:每站出现次数最多的根因;次数相同时取名称排序靠前的(与 pandas 的 mode 结果一致)
@@ -441,11 +817,12 @@ def site_stats(df):
top_cause = cause_counts.drop_duplicates('ne_id').set_index('ne_id')['root_cause'] top_cause = cause_counts.drop_duplicates('ne_id').set_index('ne_id')['root_cause']
stats['top_cause'] = stats['ne_id'].map(top_cause).fillna('') stats['top_cause'] = stats['ne_id'].map(top_cause).fillna('')
months = max(df['month'].nunique(), 1) months = max(months, 1)
stats['long'] = stats['long_count'] / months >= 1 stats['long'] = stats['long_count'] / months >= 1
stats['repeat'] = stats['count'] / months >= stats['freq_count'].fillna(4) stats['repeat'] = stats['count'] / months >= stats['freq_count'].fillna(4)
stats['total_min'] = stats['total_min'].round(1) stats['total_min'] = stats['total_min'].round(1)
stats['avg_min'] = stats['avg_min'].round(1) stats['avg_min'] = stats['avg_min'].round(1)
stats['user_h'] = stats['user_h'].round(1)
return stats.sort_values('total_min', ascending=False) return stats.sort_values('total_min', ascending=False)
@@ -453,3 +830,87 @@ def advice(cause, level):
"""处置建议文本:响应时限(按断站分级)+ 该根因的三条处置措施。""" """处置建议文本:响应时限(按断站分级)+ 该根因的三条处置措施。"""
measures = ADVICE.get(cause or '其他', ADVICE['其他']) measures = ADVICE.get(cause or '其他', ADVICE['其他'])
return f"【{LEVEL_TIP.get(level, '')}】" + ';'.join(measures) return f"【{LEVEL_TIP.get(level, '')}】" + ';'.join(measures)
def effect(filters=None, split=None):
"""
治理成效评估:以 split 日期为界,把统计周期分成治理前、治理后两段,比较两段的考核指标与问题站点变化。
- 两段各按自己的天数折算(退服率按各自时长,“月均”按 30.44 天折算),不同长度的两段可以直接比较;
- 问题站点 = 治理前被判为超长或高频的站点,逐站比较前后的月均退服次数与时长,变化超过 ±10% 记为改善 / 恶化。
split 不传时取统计周期中点所在月的 1 日。月份 / 日期筛选用来限定整个评估周期。
"""
filters = filters or {}
lo, hi = period_bounds(filters)
if split:
split = pd.Timestamp(split)
else:
split = (lo + (hi - lo) / 2).to_period('M').start_time
if not lo < split < hi:
raise ValueError(f'分界日期需在评估周期 {lo:%Y-%m-%d} ~ {hi - pd.Timedelta(days=1):%Y-%m-%d} 之内')
df = frame(filters)
n_sites = site_count(filters)
parts = {'before': (df[df['start'] < split], lo, split), 'after': (df[df['start'] >= split], split, hi)}
summary = {}
monthly = {}
for name, (part, begin, end) in parts.items():
minutes = (end - begin).total_seconds() / 60
month_equivalent = minutes / 1440 / 30.44
k = kpis(part, n_sites, minutes, max(round(month_equivalent), 1))
k['monthly_count'] = round(k['count'] / month_equivalent, 1)
k['monthly_user_h'] = round(k['user_h'] / month_equivalent)
k['days'] = round(minutes / 1440)
summary[name] = k
monthly[name] = month_equivalent
# 指标名, 字段, 越小越好
metrics = [('月均退服次数', 'monthly_count', True), ('退服率(%)', 'outage_rate', True),
('平均退服时长(分)', 'avg_min', True), ('超长退服占比(%)', 'long_ratio', True),
('重复退服率(%)', 'repeat_rate', True), ('月均受影响用户·小时', 'monthly_user_h', True),
('及时响应率(%)', 'response_rate', False), ('平均闭环时长(h)', 'avg_close_h', True)]
rows = []
for title, key, lower_is_better in metrics:
before, after = summary['before'][key], summary['after'][key]
change = round((after - before) / before * 100, 1) if before and after is not None else None
rows.append(dict(metric=title, before=before, after=after, change=change,
verdict=judge(change, lower_is_better)))
# 问题站点逐站对比
before_part = parts['before'][0]
before_stats = site_stats(before_part, max(round(monthly['before']), 1))
problem = before_stats[before_stats['long'] | before_stats['repeat']]
after_part = parts['after'][0]
after_count = after_part.groupby('ne_id').size()
after_minutes = after_part.groupby('ne_id')['duration_min'].sum()
sites = []
for site in problem.itertuples():
count_before = site.count / monthly['before']
count_after = after_count.get(site.ne_id, 0) / monthly['after']
minutes_before = site.total_min / monthly['before']
minutes_after = after_minutes.get(site.ne_id, 0) / monthly['after']
change = round((count_after - count_before) / count_before * 100, 1) if count_before else None
sites.append(dict(
ne_id=site.ne_id, site_name=site.site_name, scene=site.scene, site_type=site.site_type,
kind='超长+高频' if site.long and site.repeat else ('超长退服' if site.long else '高频重复'),
top_cause=site.top_cause, count_before=round(count_before, 2), count_after=round(count_after, 2),
min_before=round(minutes_before, 1), min_after=round(minutes_after, 1), change=change,
verdict=judge(change, True)))
sites.sort(key=lambda r: r['min_before'], reverse=True)
verdicts = pd.Series([r['verdict'] for r in sites], dtype=object)
return dict(
split=f'{split:%Y-%m-%d}', begin=f'{lo:%Y-%m-%d}', end=f'{hi - pd.Timedelta(days=1):%Y-%m-%d}',
days_before=summary['before']['days'], days_after=summary['after']['days'], rows=rows, sites=sites,
improved=int((verdicts == '改善').sum()), worsened=int((verdicts == '恶化').sum()),
flat=int((verdicts == '持平').sum()),
)
def judge(change, lower_is_better):
"""变化幅度 → 改善 / 恶化 / 持平。"""
if change is None:
return '—'
if abs(change) <= EFFECT_TOL:
return '持平'
better = change < 0 if lower_is_better else change > 0
return '改善' if better else '恶化'
+6001 -6001
View File
File diff suppressed because it is too large. Load diff
+27 -2
View File
@@ -4,7 +4,8 @@
================================================================ ================================================================
字段口径与 data.py 完全一致: 字段口径与 data.py 完全一致:
data_outage.csv 退服记录 record_id, ne_id, cell_id, district, vendor, start_time, end_time, data_outage.csv 退服记录 record_id, ne_id, cell_id, district, vendor, start_time, end_time,
duration_min, cause_category, cause_detail, cause_code, ticket_id, closed duration_min, cause_category, cause_detail, cause_code, ticket_id, closed,
handle_time, close_time(后两列为本版新增:接单时间、闭环时间)
data_alarm.csv 告警流 alarm_id, ne_id, alarm_code, alarm_name, alarm_level, category, data_alarm.csv 告警流 alarm_id, ne_id, alarm_code, alarm_name, alarm_level, category,
start_time, clear_time, duration_s, cleared, parent_alarm_id start_time, clear_time, duration_s, cleared, parent_alarm_id
sites.csv 站点库 ne_id, site_name, district, scene, site_type, vendor, cells, lng, lat sites.csv 站点库 ne_id, site_name, district, scene, site_type, vendor, cells, lng, lat
@@ -84,7 +85,7 @@ MONTH_W = [0.8, 0.75, 0.9, 1.0, 1.15, 1.35, 1.45, 1.4, 1.1] # 各月退服权
FMT = '%Y-%m-%d %H:%M:%S' FMT = '%Y-%m-%d %H:%M:%S'
OUTAGE_HEAD = ['record_id', 'ne_id', 'cell_id', 'district', 'vendor', 'start_time', 'end_time', 'duration_min', OUTAGE_HEAD = ['record_id', 'ne_id', 'cell_id', 'district', 'vendor', 'start_time', 'end_time', 'duration_min',
'cause_category', 'cause_detail', 'cause_code', 'ticket_id', 'closed'] 'cause_category', 'cause_detail', 'cause_code', 'ticket_id', 'closed', 'handle_time', 'close_time']
ALARM_HEAD = ['alarm_id', 'ne_id', 'alarm_code', 'alarm_name', 'alarm_level', 'category', 'start_time', ALARM_HEAD = ['alarm_id', 'ne_id', 'alarm_code', 'alarm_name', 'alarm_level', 'category', 'start_time',
'clear_time', 'duration_s', 'cleared', 'parent_alarm_id'] 'clear_time', 'duration_s', 'cleared', 'parent_alarm_id']
SITE_HEAD = ['ne_id', 'site_name', 'district', 'scene', 'site_type', 'vendor', 'cells', 'lng', 'lat'] SITE_HEAD = ['ne_id', 'site_name', 'district', 'scene', 'site_type', 'vendor', 'cells', 'lng', 'lat']
@@ -249,6 +250,30 @@ def generate(n=6000, seed=2026, out='.'):
noise_end = noise_start + datetime.timedelta(minutes=rng.randint(2, 30)) noise_end = noise_start + datetime.timedelta(minutes=rng.randint(2, 30))
add_alarm(rng, alarms, ne_id, noise_category, noise_start, noise_end) add_alarm(rng, alarms, ne_id, noise_category, noise_start, noise_end)
# 闭环过程时间:用独立的随机数生成器,不改变上面已有数据的随机序列。
# 接单(响应)时间 = 退服开始 + 响应时延:75% 在 1.5 小时内,18% 在 1.5~5 小时,7% 超过 5 小时;
# 已闭环的闭环时间 = max(恢复时间, 接单时间) + 10 分钟 ~ 10 小时(复核、回单);
# 未闭环的一半已接单(处理中),一半未接单(待处理)。
flow_rng = random.Random(seed + 1018)
for outage in outages:
start = datetime.datetime.strptime(outage['start_time'], FMT)
end = datetime.datetime.strptime(outage['end_time'], FMT)
r = flow_rng.random()
if r < .75:
delay = flow_rng.randint(3, 90)
elif r < .93:
delay = flow_rng.randint(91, 300)
else:
delay = flow_rng.randint(301, 1440)
handle = start + datetime.timedelta(minutes=delay)
if outage['closed'] == '是':
close = max(end, handle) + datetime.timedelta(minutes=flow_rng.randint(10, 600))
outage['handle_time'] = handle.strftime(FMT)
outage['close_time'] = close.strftime(FMT)
else:
outage['handle_time'] = handle.strftime(FMT) if flow_rng.random() < .5 else ''
outage['close_time'] = ''
# 背景告警:与退服无关的日常闪断告警 # 背景告警:与退服无关的日常闪断告警
for _ in range(n): for _ in range(n):
site = rng.choice(sites) site = rng.choice(sites)
+23 -3
View File
@@ -11,6 +11,7 @@
""" """
import io import io
import numpy as np
import pandas as pd import pandas as pd
from core import CAUSES, TIME_FMT, query from core import CAUSES, TIME_FMT, query
@@ -119,15 +120,19 @@ OUTAGE_ALIASES = {
'end_time': ['退服恢复时间', '恢复时间', '结束时间'], 'duration_min': ['退服时长', '退服时长(分钟)'], 'end_time': ['退服恢复时间', '恢复时间', '结束时间'], 'duration_min': ['退服时长', '退服时长(分钟)'],
'cause_category': ['原因大类', '退服原因'], 'cause_detail': ['原因细项'], 'cause_code': ['原因码'], 'cause_category': ['原因大类', '退服原因'], 'cause_detail': ['原因细项'], 'cause_code': ['原因码'],
'ticket_id': ['工单号'], 'closed': ['是否闭环'], 'ticket_id': ['工单号'], 'closed': ['是否闭环'],
'handled_at': ['handle_time', '接单时间', '响应时间'], 'closed_at': ['close_time', '闭环时间', '回单时间'],
} }
OUTAGE_SAVE_COLUMNS = ['record_id', 'ne_id', 'cell_id', 'start_time', 'end_time', 'duration_min', 'reported_min', OUTAGE_SAVE_COLUMNS = ['record_id', 'ne_id', 'cell_id', 'start_time', 'end_time', 'duration_min', 'reported_min',
'cause_category', 'cause_detail', 'cause_code', 'ticket_id', 'status'] 'cause_category', 'cause_detail', 'cause_code', 'ticket_id', 'status', 'handled_at',
'closed_at']
def validate_outages(df): def validate_outages(df):
""" """
退服流水校验:必填字段、站号是否在站点库、时间格式、恢复早于开始、超过 30 天、文件内及库内重复。 退服流水校验:必填字段、站号是否在站点库、时间格式、恢复早于开始、超过 30 天、文件内及库内重复、
接单 / 闭环时间格式及是否早于退服开始。
退服时长一律按“恢复时间 − 开始时间”重算;文件里人工上报的时长保存为 reported_min,供核对差错。 退服时长一律按“恢复时间 − 开始时间”重算;文件里人工上报的时长保存为 reported_min,供核对差错。
闭环状态:是否闭环=是 或有闭环时间 → 已闭环;否则有接单时间 → 处理中;其余 → 待处理。
""" """
df = rename_columns(df, OUTAGE_ALIASES) df = rename_columns(df, OUTAGE_ALIASES)
@@ -153,6 +158,10 @@ def validate_outages(df):
start_text = clean_text(df['start_time']) # 问题行里原样显示用户填写的开始时间 start_text = clean_text(df['start_time']) # 问题行里原样显示用户填写的开始时间
start = parse_time(df['start_time']) start = parse_time(df['start_time'])
end = parse_time(df['end_time']) end = parse_time(df['end_time'])
handled = parse_time(df['handled_at'])
closed = parse_time(df['closed_at'])
handled_given = clean_text(df['handled_at']) != ''
closed_given = clean_text(df['closed_at']) != ''
known_sites = {row['ne_id'] for row in query('SELECT ne_id FROM sites')} known_sites = {row['ne_id'] for row in query('SELECT ne_id FROM sites')}
existing = {(row['ne_id'], row['start_time']) for row in query('SELECT ne_id, start_time FROM outages')} existing = {(row['ne_id'], row['start_time']) for row in query('SELECT ne_id, start_time FROM outages')}
@@ -182,6 +191,14 @@ def validate_outages(df):
elif key in seen_in_file or key in existing: elif key in seen_in_file or key in existing:
reasons.append('重复记录(同站同开始时间)') reasons.append('重复记录(同站同开始时间)')
seen_in_file.add(key) seen_in_file.add(key)
if handled_given[i] and pd.isna(handled[i]):
reasons.append('接单时间格式错误')
elif handled_given[i] and handled[i] < start[i]:
reasons.append('接单时间早于退服开始')
if closed_given[i] and pd.isna(closed[i]):
reasons.append('闭环时间格式错误')
elif closed_given[i] and closed[i] < start[i]:
reasons.append('闭环时间早于退服开始')
if reasons: if reasons:
errors.append({'row': i + 2, 'key': ne_id, 'start_time': start_text[i], 'reason': ';'.join(reasons)}) errors.append({'row': i + 2, 'key': ne_id, 'start_time': start_text[i], 'reason': ';'.join(reasons)})
@@ -195,7 +212,10 @@ def validate_outages(df):
good['end_time'] = good_end.dt.strftime(TIME_FMT) good['end_time'] = good_end.dt.strftime(TIME_FMT)
good['reported_min'] = pd.to_numeric(good['duration_min'], errors='coerce').astype(float) good['reported_min'] = pd.to_numeric(good['duration_min'], errors='coerce').astype(float)
good['duration_min'] = ((good_end - good_start).dt.total_seconds() / 60).round(1) good['duration_min'] = ((good_end - good_start).dt.total_seconds() / 60).round(1)
good['status'] = good['closed'].map({'是': '已闭环', '否': '待处理'}).fillna('待处理') good['handled_at'] = handled.loc[good_rows].dt.strftime(TIME_FMT)
good['closed_at'] = closed.loc[good_rows].dt.strftime(TIME_FMT)
is_closed = (clean_text(good['closed']) == '是') | good['closed_at'].notna()
good['status'] = np.select([is_closed, good['handled_at'].notna()], ['已闭环', '处理中'], default='待处理')
# 原因大类只接受系统定义的 7 类,其他写法视为未填写 # 原因大类只接受系统定义的 7 类,其他写法视为未填写
good['cause_category'] = good['cause_category'].where(good['cause_category'].isin(CAUSES), None) good['cause_category'] = good['cause_category'].where(good['cause_category'].isin(CAUSES), None)
+9 -1
View File
@@ -1,6 +1,6 @@
/* 内容层组件。只放有实际逻辑的组件;纯布局直接写 className(.grid / .row / .grow / .note / .chips,见 app.css)。 */ /* 内容层组件。只放有实际逻辑的组件;纯布局直接写 className(.grid / .row / .grow / .note / .chips,见 app.css)。 */
import {Card, Fragment, html, Picker, useState} from 'vendor'; import {Card, Fragment, html, Picker, useState} from 'vendor';
import {CAUSE_COLOR, LEVEL_TONE, STATUS_TONE} from '../format.js'; import {CAUSE_COLOR, LEVEL_TONE, SLA_TONE, STATUS_TONE} from '../format.js';
/** 内容卡片:不透明,可带标题栏。span 为 12 栅格中占的列数 */ /** 内容卡片:不透明,可带标题栏。span 为 12 栅格中占的列数 */
export function Section({title, hint, extra, span = 12, children}) { export function Section({title, hint, extra, span = 12, children}) {
@@ -40,6 +40,14 @@ export function StatusTag({value}) {
return html`<span className="tag" data-tone=${STATUS_TONE[value]}>${dot}${value}</span>`; return html`<span className="tag" data-tone=${STATUS_TONE[value]}>${dot}${value}</span>`;
} }
/** 响应时限判定标签(及时 / 超时 / 超时未响应 / 未到期 / 无记录) */
export const SlaTag = ({value}) => value && html`<span className="tag" data-tone=${SLA_TONE[value]}>${value}</span>`;
/** 变化判定:改善绿、恶化红、持平灰 */
export const VerdictTag = ({value}) => value && value !== '—'
? html`<span className="tag" data-tone=${{改善: 'green', 恶化: 'red', 持平: 'gray'}[value]}>${value}</span>`
: '—';
/** 根因:色点 + 名称 */ /** 根因:色点 + 名称 */
export const Cause = ({value}) => value && html` export const Cause = ({value}) => value && html`
<span className="cause"><i style=${{background: CAUSE_COLOR[value] || '#8e8e93'}}></i>${value}</span>`; <span className="cause"><i style=${{background: CAUSE_COLOR[value] || '#8e8e93'}}></i>${value}</span>`;
+2
View File
@@ -56,5 +56,7 @@ export const BLUE_SCALE = ['#e8f2ff', '#a9cfff', '#4da3ff', '#0062cc'];
/** 断站分级、闭环状态 -> 标签色调(见 app.css 中的 .tag[data-tone]) */ /** 断站分级、闭环状态 -> 标签色调(见 app.css 中的 .tag[data-tone]) */
export const LEVEL_TONE = {重要: 'red', 主要: 'orange', 次要: 'gray'}; export const LEVEL_TONE = {重要: 'red', 主要: 'orange', 次要: 'gray'};
export const STATUS_TONE = {已闭环: 'green', 处理中: 'blue', 待处理: 'yellow'}; export const STATUS_TONE = {已闭环: 'green', 处理中: 'blue', 待处理: 'yellow'};
/** 响应时限判定 -> 标签色调 */
export const SLA_TONE = {及时: 'green', 超时: 'orange', 超时未响应: 'red', 未到期: 'blue', 无记录: 'gray'};
export const ROLE_NAME = {admin: '管理员', user: '普通用户'}; export const ROLE_NAME = {admin: '管理员', user: '普通用户'};
+80 -9
View File
@@ -1,15 +1,34 @@
import {html} from 'vendor'; import {html, TextField, useState} from 'vendor';
import {useApp, useFetch} from '../api.js'; import {useApp, useFetch} from '../api.js';
import {accentMax, fixed, num} from '../format.js'; import {accentMax, fixed, num} from '../format.js';
import {EChart} from '../components/EChart.js'; import {EChart} from '../components/EChart.js';
import {DataTable} from '../components/DataTable.js'; import {DataTable} from '../components/DataTable.js';
import {Kpi, Loading, Section} from '../components/ui.js'; import {Cause, Kpi, Loading, Section, VerdictTag} from '../components/ui.js';
const BENCH_COLUMNS = [ const BENCH_COLUMNS = [
{key: 'scene', title: '场景'}, {key: 'sites', title: '站点数'}, {key: 'count', title: '退服次数'}, {key: 'scene', title: '场景'}, {key: 'sites', title: '站点数'}, {key: 'count', title: '退服次数'},
{key: 'total_h', title: '累计时长(h)'}, {key: 'avg_min', title: '平均时长(min)'}, {key: 'outage_rate', title: '退服率(%)'}, {key: 'total_h', title: '累计时长(h)'}, {key: 'avg_min', title: '平均时长(min)'}, {key: 'outage_rate', title: '退服率(%)'},
{key: 'availability', title: '可用率(%)'}, {key: 'long_ratio', title: '超长占比(%)'}, {key: 'repeat_rate', title: '重复退服率(%)'}, {key: 'availability', title: '可用率(%)'}, {key: 'long_ratio', title: '超长占比(%)'}, {key: 'repeat_rate', title: '重复退服率(%)'},
{key: 'users', title: '受影响用户', render: num}, {key: 'loss_gb', title: '业务损失(GB)', render: num}, {key: 'users', title: '受影响用户', render: num}, {key: 'user_h', title: '用户·小时', render: num},
{key: 'loss_gb', title: '业务损失(GB)', render: num},
{key: 'response_rate', title: '及时响应率(%)', render: v => v ?? '—'},
{key: 'avg_close_h', title: '平均闭环(h)', render: v => v ?? '—'},
];
const EFFECT_COLUMNS = [
{key: 'metric', title: '指标'}, {key: 'before', title: '治理前', render: v => (v == null ? '—' : num(v))},
{key: 'after', title: '治理后', render: v => (v == null ? '—' : num(v))},
{key: 'change', title: '变化(%)', render: v => (v === null ? '—' : `${v > 0 ? '+' : ''}${v}`)},
{key: 'verdict', title: '判定', render: v => html`<${VerdictTag} value=${v}/>`},
];
const EFFECT_SITE_COLUMNS = [
{key: 'site_name', title: '基站名称'}, {key: 'scene', title: '场景'}, {key: 'site_type', title: '站型'},
{key: 'kind', title: '治理前问题'}, {key: 'top_cause', title: '主要根因', render: v => html`<${Cause} value=${v}/>`},
{key: 'count_before', title: '月均次数·前'}, {key: 'count_after', title: '月均次数·后'},
{key: 'min_before', title: '月均时长·前(分)', render: num}, {key: 'min_after', title: '月均时长·后(分)', render: num},
{key: 'change', title: '次数变化(%)', render: v => (v === null ? '—' : `${v > 0 ? '+' : ''}${v}`)},
{key: 'verdict', title: '判定', render: v => html`<${VerdictTag} value=${v}/>`},
]; ];
/** 场景柱状图配置:最大的场景用强调色标出 */ /** 场景柱状图配置:最大的场景用强调色标出 */
@@ -24,14 +43,15 @@ export default function Impact() {
const {filters} = useApp(); const {filters} = useApp();
const {data} = useFetch('/api/dashboard', filters); const {data} = useFetch('/api/dashboard', filters);
if (!data) return html`<${Loading}/>`; if (!data) return html`<${Loading}/>`;
const {kpi: k, bench} = data; const {kpi: k, bench, period} = data;
return html` return html`
<div className="grid"> <div className="grid">
<${Section} title="影响评估"> <${Section} title="影响评估" hint=${`统计周期 ${period.days} 天(${period.months} 个月)· ${num(k.sites)} 个站`}>
<div className="kpis"> <div className="kpis">
<${Kpi} title="受影响用户" value=${num(k.users)} unit="人次" sub="小区数 × 单小区在线用户 × 忙闲时系数"/> <${Kpi} title="受影响用户" value=${num(k.users)} unit="人次" sub="小区数 × 单小区在线用户 × 忙闲时 × 覆盖补偿"/>
<${Kpi} title="业务损失" value=${num(k.loss_gb)} unit="GB" sub="小区话务 × 退服时长"/> <${Kpi} title="受影响用户·小时" value=${num(k.user_h)} unit="人·h" sub="受影响用户 × 退服时长"/>
<${Kpi} title="业务损失" value=${num(k.loss_gb)} unit="GB" sub="小区话务 × 忙闲时 × 覆盖补偿 × 退服时长"/>
<${Kpi} title="退服率" value=${fixed(k.outage_rate)} unit="%"/> <${Kpi} title="退服率" value=${fixed(k.outage_rate)} unit="%"/>
<${Kpi} title="网络可用率" value=${fixed(k.availability)} unit="%"/> <${Kpi} title="网络可用率" value=${fixed(k.availability)} unit="%"/>
<${Kpi} title="平均退服时长" value=${k.avg_min} unit="min"/> <${Kpi} title="平均退服时长" value=${k.avg_min} unit="min"/>
@@ -39,14 +59,65 @@ export default function Impact() {
<${Kpi} title="重复退服率" value=${fixed(k.repeat_rate)} unit="%"/> <${Kpi} title="重复退服率" value=${fixed(k.repeat_rate)} unit="%"/>
</div> </div>
</${Section}> </${Section}>
<${Section} title="闭环管控" hint="响应时限从退服开始算:重要 2 小时、主要 4 小时、次要 24 小时">
<div className="kpis">
<${Kpi} title="闭环率" value=${fixed(k.closed)} unit="%"/>
<${Kpi} title="及时响应率" value=${fixed(k.response_rate) || '—'} unit="%" sub="及时 ÷(及时 + 超时 + 超时未响应)"/>
<${Kpi} title="平均闭环时长" value=${k.avg_close_h ?? '—'} unit="h" sub="退服开始 → 闭环;无接单 / 闭环时间的记录不计"/>
<${Kpi} title="超时未闭环" value=${num(k.overdue_open)} unit="条" sub="可在退服流水管理中筛选处理"/>
</div>
</${Section}>
<${Section} span=${6} title="分场景退服率" hint="单位:%"> <${Section} span=${6} title="分场景退服率" hint="单位:%">
<${EChart} option=${sceneBar(bench, bench.map(x => x.outage_rate))}/> <${EChart} option=${sceneBar(bench, bench.map(x => x.outage_rate))}/>
</${Section}> </${Section}>
<${Section} span=${6} title="分场景受影响用户" hint="单位:万人次"> <${Section} span=${6} title="分场景受影响用户·小时" hint="单位:万人·h">
<${EChart} option=${sceneBar(bench, bench.map(x => +(x.users / 1e4).toFixed(1)))}/> <${EChart} option=${sceneBar(bench, bench.map(x => +(x.user_h / 1e4).toFixed(1)))}/>
</${Section}> </${Section}>
<${Section} title="通信场景运维考核指标横向对标"> <${Section} title="通信场景运维考核指标横向对标">
<${DataTable} rowKey="scene" columns=${BENCH_COLUMNS} rows=${bench}/> <${DataTable} rowKey="scene" columns=${BENCH_COLUMNS} rows=${bench}/>
</${Section}> </${Section}>
<${Effect}/>
<${Section} title="口径说明">
<div className="note">
退服率 = 总退服时长 ÷(站点数 × 统计周期);统计周期取顶栏的月份 / 日期范围,未筛选时为全部流水首日至末日。
重复退服率 = 高频重复站点数 ÷ 站点数。“月均”按统计周期的自然月数计算。
覆盖补偿:站点 500 米内每有一个邻站,受影响用户与话务损失下降 10%(室分站 5%),最低保留 40%。
</div>
</${Section}>
</div>`; </div>`;
} }
/** 治理成效评估:以分界日期把统计周期分成前后两段,比较考核指标与问题站点的变化 */
function Effect() {
const {filters} = useApp();
const [split, setSplit] = useState('');
const {data} = useFetch('/api/effect', {...filters, split});
const picker = html`
<${TextField} type="date" label="治理分界日期" labelHidden controlSize="small" value=${split || data?.split || ''}
onChange=${e => setSplit(e.target.value)}/>`;
if (!data) return html`<${Section} title="治理成效评估" extra=${picker}><${Loading}/></${Section}>`;
const hint = `${data.begin} ~ ${data.split}(${data.days_before} 天)对比 ${data.split} ~ ${data.end}(${data.days_after} 天)`;
return html`
<${Section} title="治理成效评估" hint=${hint} extra=${picker}>
<div className="grid">
<div className="span-5">
<${DataTable} rowKey="metric" columns=${EFFECT_COLUMNS} rows=${data.rows}/>
<div className="note">
两段各按自身天数折算,长度不同也可直接比较;变化在 ±10% 以内记为持平。
退服存在季节性(夏季雷雨、高温停电多),跨季节对比时请结合同期数据判断。
</div>
</div>
<div className="span-7">
<div className="chips">
<div className="chip"><span>治理前问题站点</span>${data.sites.length} 个</div>
<div className="chip" data-tone="green"><span>改善</span>${data.improved}</div>
<div className="chip"><span>持平</span>${data.flat}</div>
<div className="chip"><span>恶化</span>${data.worsened}</div>
</div>
<${DataTable} rowKey="ne_id" columns=${EFFECT_SITE_COLUMNS} rows=${data.sites} maxHeight=${420}/>
</div>
</div>
</${Section}>`;
}
+11 -1
View File
@@ -2,7 +2,7 @@ import {Fragment, GlassButton, GlassCheckbox, html, SearchField, TextField, useS
import {api, useApp, useFetch} from '../api.js'; import {api, useApp, useFetch} from '../api.js';
import {DataTable, Pager} from '../components/DataTable.js'; import {DataTable, Pager} from '../components/DataTable.js';
import {FormDialog} from '../components/FormDialog.js'; import {FormDialog} from '../components/FormDialog.js';
import {Cause, LevelTag, Loading, Section, Select, StatusTag} from '../components/ui.js'; import {Cause, LevelTag, Loading, Section, Select, SlaTag, StatusTag} from '../components/ui.js';
const PAGE_SIZE = 20; const PAGE_SIZE = 20;
@@ -60,6 +60,10 @@ export default function Outages() {
{key: 'root_cause', title: '判定根因', render: v => html`<${Cause} value=${v}/>`}, {key: 'root_cause', title: '判定根因', render: v => html`<${Cause} value=${v}/>`},
{key: 'level', title: '分级', render: v => html`<${LevelTag} value=${v}/>`}, {key: 'level', title: '分级', render: v => html`<${LevelTag} value=${v}/>`},
{key: 'status', title: '状态', render: v => html`<${StatusTag} value=${v}/>`}, {key: 'status', title: '状态', render: v => html`<${StatusTag} value=${v}/>`},
{key: 'sla', title: '响应时限', render: (v, r) => html`
<span title=${r.response_min !== '' ? `响应 ${r.response_min} 分钟,时限 ${r.limit_h} 小时` : `时限 ${r.limit_h} 小时`}>
<${SlaTag} value=${v}/></span>`},
{key: 'response_min', title: '响应(分)', sortable: true},
{key: 'affected_users', title: '影响用户', sortable: true}, {key: 'affected_users', title: '影响用户', sortable: true},
{key: 'id', title: '操作', render: (_, r) => html` {key: 'id', title: '操作', render: (_, r) => html`
<div className="row nowrap"> <div className="row nowrap">
@@ -78,6 +82,7 @@ export default function Outages() {
<${Select} label="闭环状态" all="全部闭环状态" options=${meta.status} value=${draft.status} onChange=${setD('status')}/> <${Select} label="闭环状态" all="全部闭环状态" options=${meta.status} value=${draft.status} onChange=${setD('status')}/>
<${GlassCheckbox} label="仅超长" checked=${!!draft.long} onCheckedChange=${v => setD('long')(v ? 1 : '')}/> <${GlassCheckbox} label="仅超长" checked=${!!draft.long} onCheckedChange=${v => setD('long')(v ? 1 : '')}/>
<${GlassCheckbox} label="仅时长差错" checked=${!!draft.diff} onCheckedChange=${v => setD('diff')(v ? 1 : '')}/> <${GlassCheckbox} label="仅时长差错" checked=${!!draft.diff} onCheckedChange=${v => setD('diff')(v ? 1 : '')}/>
<${GlassCheckbox} label="仅超时未闭环" checked=${!!draft.overdue} onCheckedChange=${v => setD('overdue')(v ? 1 : '')}/>
<${GlassButton} variant="glassProminent" onClick=${search}>查询</${GlassButton}> <${GlassButton} variant="glassProminent" onClick=${search}>查询</${GlassButton}>
<span className="grow"></span> <span className="grow"></span>
${isAdmin && html` ${isAdmin && html`
@@ -151,6 +156,11 @@ function OutageDialog({row, onClose, onSaved}) {
<${Select} name="status" label="闭环状态" showLabel options=${meta.status} value=${row.status || '待处理'}/> <${Select} name="status" label="闭环状态" showLabel options=${meta.status} value=${row.status || '待处理'}/>
<${TextField} name="handler" label="处理人" defaultValue=${row.handler}/> <${TextField} name="handler" label="处理人" defaultValue=${row.handler}/>
<${TextField} className="full" name="remark" label="处理说明" multiline rows=${2} defaultValue=${row.remark}/> <${TextField} className="full" name="remark" label="处理说明" multiline rows=${2} defaultValue=${row.remark}/>
${!isNew && html`<div className="note full">
接单时间:${row.handled_at || '—'} 闭环时间:${row.closed_at || '—'}
 响应时限:${row.limit_h} 小时(${row.level}) 判定:${row.sla}<br/>
状态改为“处理中 / 已闭环”时自动记录接单时间,改为“已闭环”时自动记录闭环时间。
</div>`}
${row.advice && html`<div className="note full"><b>系统判定根因:${row.root_cause}</b>(${row.evidence})<br/>${row.advice}</div>`} ${row.advice && html`<div className="note full"><b>系统判定根因:${row.root_cause}</b>(${row.evidence})<br/>${row.advice}</div>`}
</${FormDialog}>`; </${FormDialog}>`;
} }
+34 -7
View File
@@ -7,11 +7,23 @@ import {Cause, LevelTag, Loading, Section, Select, StatusTag} from '../component
const PAGE_SIZE = 15; const PAGE_SIZE = 15;
const WINDOW_COLUMNS = [
{key: 'priority', title: '优先级'}, {key: 'category', title: '告警类别', render: v => html`<${Cause} value=${v}/>`},
{key: 'default_min', title: '默认'}, {key: 'window_min', title: '当前'}, {key: 'samples', title: '标定样本', render: num},
{key: 'source', title: '来源'},
];
const ASSOC_COLUMNS = [
{key: 'category', title: '告警类别', render: v => html`<${Cause} value=${v}/>`}, {key: 'window_min', title: '窗口(分)'},
{key: 'alarms', title: '告警数', render: num}, {key: 'hits', title: '其后退服', render: num},
{key: 'support', title: '支持度(%)'}, {key: 'confidence', title: '置信度(%)'}, {key: 'lift', title: '提升度'},
];
export default function RootCause() { export default function RootCause() {
const {filters, isAdmin} = useApp(); const {filters, isAdmin} = useApp();
const toast = useToast(); const toast = useToast();
const {data: confusion} = useFetch('/api/confusion', filters); const {data: confusion} = useFetch('/api/confusion', filters);
const {data: stats} = useFetch('/api/stats', {...filters, dim: 'root_cause'}); const {data: stats} = useFetch('/api/stats', {...filters, dim: 'root_cause'});
const {data: assoc, reload: reloadAssoc} = useFetch('/api/association', filters);
const [lastRun, setLastRun] = useState(null); const [lastRun, setLastRun] = useState(null);
const [running, setRunning] = useState(false); const [running, setRunning] = useState(false);
@@ -20,6 +32,7 @@ export default function RootCause() {
try { try {
const r = await api('/api/analyze', {method: 'POST'}); const r = await api('/api/analyze', {method: 'POST'});
setLastRun(r); setLastRun(r);
reloadAssoc();
toast({message: `全量分析完成:${num(r.n)} 条,耗时 ${r.elapsed}s,准确率 ${r.acc}%`, tone: 'success'}); toast({message: `全量分析完成:${num(r.n)} 条,耗时 ${r.elapsed}s,准确率 ${r.acc}%`, tone: 'success'});
} catch (e) { } catch (e) {
toast({message: e.message, tone: 'error'}); toast({message: e.message, tone: 'error'});
@@ -27,7 +40,7 @@ export default function RootCause() {
setRunning(false); setRunning(false);
}; };
if (!confusion || !stats) return html`<${Loading}/>`; if (!confusion || !stats || !assoc) return html`<${Loading}/>`;
// 交叉矩阵:matrix[i][j] = 上报原因 i、系统判定 j 的条数;对角线 i === j 为判定一致 // 交叉矩阵:matrix[i][j] = 上报原因 i、系统判定 j 的条数;对角线 i === j 为判定一致
const {causes, matrix} = confusion; const {causes, matrix} = confusion;
@@ -75,19 +88,32 @@ export default function RootCause() {
<div className="grid"> <div className="grid">
<${Section} span=${5} title="根因判定规则" extra=${runButton}> <${Section} span=${5} title="根因判定规则" extra=${runButton}>
<div className="note"> <div className="note">
<b>第一级 · 告警衍生关系</b>:“小区退服”告警若带 parent_alarm_id,直接取父告警类别作为根因。<br/> <b>预处理 · 衍生告警归并</b>:带父告警的告警按父告警类别计(停电引起的传输告警归为电源故障)。<br/>
<b>第二级 · 时间窗关联</b>:回溯同站在退服前仍未清除的告警,按优先级取首个命中:<br/> <b>第一级 · 衍生关系(置信度高)</b>:与退服开始相差 ≤ 2 分钟的“小区退服”告警带父告警时,取父告警类别。<br/>
① 割接施工 30 分钟 ② 电源故障 300 分钟(考虑蓄电池后备)③ 机房环境 180 分钟(考虑过热保护延时)<br/> <b>第二级 · 时间窗(置信度中)</b>:同站在退服时仍未清除、提前量在该类时间窗内的告警。<br/>
④ 传输中断 15 分钟 ⑤ 天馈故障 30 分钟 ⑥ 设备硬件 15 分钟<br/> <b>第三级 · 疑似(置信度低)</b>:时间窗放宽到 3 倍才命中,提示人工复核。<br/>
停电衍生的传输告警因优先级归为电源故障;均未命中判为“其他”,派单现场核查。 多条命中时依次比较:级别 → 是否随业务恢复清除(±15 分钟)→ 规则优先级 → 时间早。
时间窗在每次全量分析时按已闭环流水的提前量 95% 分位数重新标定(不小于默认值);均未命中判为“其他”。
</div> </div>
<div className="chips"> <div className="chips">
<div className="chip"><span>参与校验</span>${num(total)} 条</div> <div className="chip"><span>参与校验</span>${num(total)} 条</div>
<div className="chip" data-tone="green"><span>判定准确率</span>${total ? (hit / total * 100).toFixed(2) : 0}%</div> <div className="chip" data-tone="green"><span>判定准确率</span>${total ? (hit / total * 100).toFixed(2) : 0}%</div>
${confusion.by_confidence.map(c => html`
<div className="chip" key=${c.level}><span>置信度${c.level}</span>${num(c.n)} 条 / ${c.acc}%</div>`)}
<div className="chip"><span>上次分析</span>${lastRun ? `${num(lastRun.n)} 条 / ${lastRun.elapsed}s` : '—'}</div> <div className="chip"><span>上次分析</span>${lastRun ? `${num(lastRun.n)} 条 / ${lastRun.elapsed}s` : '—'}</div>
</div> </div>
</${Section}> </${Section}>
<${Section} span=${7} title="上报原因 × 系统判定根因 交叉矩阵" hint="对角线为判定一致"><${EChart} option=${matrixChart}/></${Section}> <${Section} span=${7} title="上报原因 × 系统判定根因 交叉矩阵" hint="对角线为判定一致"><${EChart} option=${matrixChart}/></${Section}>
<${Section} span=${5} title="时间窗标定" hint="单位:分钟">
<${DataTable} rowKey="category" columns=${WINDOW_COLUMNS} rows=${assoc.windows}/>
</${Section}>
<${Section} span=${7} title="告警 → 退服 关联强度" hint="提升度 > 1 表示该类告警出现后退服概率高于随机">
<${DataTable} rowKey="category" columns=${ASSOC_COLUMNS} rows=${assoc.rows}/>
<div className="note">
支持度 = 开始前 W 分钟内有该类告警的退服 ÷ 退服总数;置信度 = 其后 W 分钟内发生退服的告警 ÷ 该类告警总数;
提升度 = 置信度 ÷ 随机一个 W 分钟窗口内遇到退服的概率(W 为该类时间窗)。
</div>
</${Section}>
<${Section} title="根因维度统计" hint="柱高为平均退服时长(分钟)"><${EChart} height=${260} option=${causeChart}/></${Section}> <${Section} title="根因维度统计" hint="柱高为平均退服时长(分钟)"><${EChart} height=${260} option=${causeChart}/></${Section}>
<${AdviceList}/> <${AdviceList}/>
</div>`; </div>`;
@@ -108,7 +134,8 @@ function AdviceList() {
const columns = [ const columns = [
{key: 'ne_id', title: '基站编号'}, {key: 'site_name', title: '基站名称'}, {key: 'scene', title: '场景'}, {key: 'ne_id', title: '基站编号'}, {key: 'site_name', title: '基站名称'}, {key: 'scene', title: '场景'},
{key: 'start_time', title: '退服开始'}, {key: 'duration_min', title: '时长(分)'}, {key: 'start_time', title: '退服开始'}, {key: 'duration_min', title: '时长(分)'},
{key: 'root_cause', title: '根因', render: v => html`<${Cause} value=${v}/>`}, {key: 'evidence', title: '关联告警证据'}, {key: 'root_cause', title: '根因', render: v => html`<${Cause} value=${v}/>`}, {key: 'confidence', title: '置信度'},
{key: 'evidence', title: '关联告警证据'},
{key: 'level', title: '分级', render: v => html`<${LevelTag} value=${v}/>`}, {key: 'level', title: '分级', render: v => html`<${LevelTag} value=${v}/>`},
{key: 'status', title: '状态', render: v => html`<${StatusTag} value=${v}/>`}, {key: 'status', title: '状态', render: v => html`<${StatusTag} value=${v}/>`},
{key: 'advice', title: '处置建议', wrap: true}, {key: 'advice', title: '处置建议', wrap: true},
+21
View File
@@ -18,6 +18,7 @@ export default function Sites() {
const [keyword, setKeyword] = useState(''); const [keyword, setKeyword] = useState('');
const [editScene, setEditScene] = useState(null); const [editScene, setEditScene] = useState(null);
const [editSite, setEditSite] = useState(null); const [editSite, setEditSite] = useState(null);
const [editType, setEditType] = useState(null);
if (!data) return html`<${Loading}/>`; if (!data) return html`<${Loading}/>`;
@@ -62,6 +63,17 @@ export default function Sites() {
<${DataTable} rowKey="scene" rows=${data.scenes} columns=${[{key: 'scene', title: '场景'}, ...SCENE_FIELDS, ...editColumn(setEditScene)]}/> <${DataTable} rowKey="scene" rows=${data.scenes} columns=${[{key: 'scene', title: '场景'}, ...SCENE_FIELDS, ...editColumn(setEditScene)]}/>
</${Section}> </${Section}>
<${Section} span=${5} title="场景 × 站型 站点数量"><${EChart} option=${mixChart}/></${Section}> <${Section} span=${5} title="场景 × 站型 站点数量"><${EChart} option=${mixChart}/></${Section}>
<${Section} span=${5} title="站型阈值系数" hint="站点实际阈值 = 场景阈值 × 站型系数">
<${DataTable} rowKey="site_type" rows=${data.site_types} columns=${[
{key: 'site_type', title: '站型'}, {key: 'long_factor', title: '超长时长系数'},
{key: 'freq_factor', title: '高频次数系数'}, ...editColumn(setEditType)]}/>
</${Section}>
<${Section} span=${7} title="场景 × 站型 实际阈值" hint="超长阈值(分钟) / 高频阈值(次/月)">
<${DataTable} rowKey="scene" rows=${data.scenes} columns=${[
{key: 'scene', title: '场景'},
...data.site_types.map(t => ({key: t.site_type, title: t.site_type, render: (_, s) =>
`${+(s.long_min * t.long_factor).toFixed(1)} / ${+(s.freq_count * t.freq_factor).toFixed(1)}`}))]}/>
</${Section}>
<${Section} title="站点场景标签与站型分级" extra=${siteFilters}> <${Section} title="站点场景标签与站型分级" extra=${siteFilters}>
<${DataTable} rowKey="ne_id" rows=${sites} columns=${[ <${DataTable} rowKey="ne_id" rows=${sites} columns=${[
{key: 'ne_id', title: '基站编号'}, {key: 'site_name', title: '站点名称'}, {key: 'vendor', title: '厂家'}, {key: 'ne_id', title: '基站编号'}, {key: 'site_name', title: '站点名称'}, {key: 'vendor', title: '厂家'},
@@ -77,6 +89,15 @@ export default function Sites() {
<${TextField} key=${f.key} name=${f.key} label=${f.title} type="number" step=${f.step || '1'} min="0" required <${TextField} key=${f.key} name=${f.key} label=${f.title} type="number" step=${f.step || '1'} min="0" required
defaultValue=${editScene[f.key]}/>`)} defaultValue=${editScene[f.key]}/>`)}
</${FormDialog}>`} </${FormDialog}>`}
${editType && html`
<${FormDialog} onClose=${() => setEditType(null)} title=${`站型系数 · ${editType.site_type}`}
description="系数 < 1 表示从严,> 1 表示放宽;保存后立即生效"
onSubmit=${d => save(`/api/site_types/${editType.site_type}`, d, '站型系数已更新')}>
<${TextField} name="long_factor" label="超长时长系数" type="number" step="0.05" min="0.1" max="10" required
defaultValue=${editType.long_factor}/>
<${TextField} name="freq_factor" label="高频次数系数" type="number" step="0.05" min="0.1" max="10" required
defaultValue=${editType.freq_factor}/>
</${FormDialog}>`}
${editSite && html` ${editSite && html`
<${FormDialog} onClose=${() => setEditSite(null)} title="编辑站点分级" description=${`${editSite.ne_id} · ${editSite.site_name}`} <${FormDialog} onClose=${() => setEditSite(null)} title="编辑站点分级" description=${`${editSite.ne_id} · ${editSite.site_name}`}
onSubmit=${d => save(`/api/sites/${editSite.ne_id}`, d, '站点已更新')}> onSubmit=${d => save(`/api/sites/${editSite.ne_id}`, d, '站点已更新')}>
+174
View File
@@ -0,0 +1,174 @@
"""测试公共工具:在 tests/_work 下的临时数据库里加载数据集并调用 core 的分析逻辑,不影响 data/outage.db。"""
import os
import random
import sys
PROJECT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
sys.path.insert(0, PROJECT)
WORK = os.path.join(PROJECT, 'tests', '_work')
os.makedirs(WORK, exist_ok=True)
os.environ['OUTAGE_DB'] = os.path.join(WORK, 'app.db') # 必须在 import core 之前设置
import pandas as pd # noqa: E402
from werkzeug.security import generate_password_hash # noqa: E402
import core # noqa: E402
import data_gen # noqa: E402
import importer # noqa: E402
FAILED = []
def check(name, ok, detail=''):
"""打印一条检查结果;失败的记入 FAILED,脚本结束时以非 0 退出码返回。"""
print(f" [{'通过' if ok else '失败'}] {name} {detail}")
if not ok:
FAILED.append(name)
def finish():
print('\n全部检查通过' if not FAILED else f'\n{len(FAILED)} 项未通过:' + '、'.join(FAILED))
sys.exit(1 if FAILED else 0)
def use_db(name):
"""切换到 tests/_work/<name>.db(删除重建)。"""
path = os.path.join(WORK, f'{name}.db')
if os.path.exists(path):
os.remove(path)
core.DB_PATH = path
core.data_changed()
with core.db() as conn:
conn.executescript(core.SCHEMA)
for scene, values in core.SCENE_DEFAULT.items():
conn.execute('INSERT INTO scenes VALUES(?,?,?,?,?)', (scene, *values))
conn.execute("INSERT INTO users(username, password, role, created_at) VALUES('admin', ?, 'admin', '')",
(generate_password_hash('admin123'),)) # 测试库的接口测试账号
core.migrate() # 与正式启动一致:补站型系数默认值
def load(name, sites, alarms, outages, calibrate_until=None):
"""建库并写入 sites / alarms / outages(全文本 DataFrame),做根因分析,返回 full_frame()。
calibrate_until 指定时,时间窗只用该日期之前的流水标定(留出检验,避免用测试数据标定)。"""
use_db(name)
with core.db() as conn:
sites.to_sql('sites', conn, if_exists='append', index=False)
core.save_alarms(alarms.astype(object).where(alarms.notna(), None))
good, _ = importer.validate_outages(outages)
core.save_outages(good)
if calibrate_until:
core.calibrate_windows(until=calibrate_until)
core.analyze_all(recalibrate=False)
else:
core.analyze_all()
core.data_changed()
return core.full_frame()
def read_dir(directory):
def read(filename):
return pd.read_csv(os.path.join(directory, filename), encoding='utf-8-sig', dtype=str)
return read('sites.csv'), read('data_alarm.csv'), read('data_outage.csv')
def sample_data():
return read_dir(os.path.join(PROJECT, 'data'))
def generate(name, lead=None, rows=6000, seed=2026):
"""用 data_gen 生成一份数据到 tests/_work/gen_<name>;lead 可覆盖告警提前量分布。"""
out = os.path.join(WORK, 'gen_' + name)
saved = dict(data_gen.LEAD)
if lead is not None:
data_gen.LEAD.clear()
data_gen.LEAD.update(lead)
try:
data_gen.generate(n=rows, seed=seed, out=out)
finally:
data_gen.LEAD.clear()
data_gen.LEAD.update(saved)
return read_dir(out)
def score(df):
"""以生成器真值(cause_category)计算总体准确率与各类 precision / recall。"""
d = df[df['cause_category'].isin(core.CAUSES)]
if d.empty:
return 0.0, []
accuracy = (d['cause_category'] == d['root_cause']).mean() * 100
rows = []
for cause in core.CAUSES:
tp = ((d['cause_category'] == cause) & (d['root_cause'] == cause)).sum()
precision = tp / max((d['root_cause'] == cause).sum(), 1) * 100
recall = tp / max((d['cause_category'] == cause).sum(), 1) * 100
f1 = 2 * precision * recall / (precision + recall) if precision + recall else 0
rows.append((cause, round(precision, 1), round(recall, 1), round(f1, 1)))
return round(accuracy, 2), rows
# ---------------------------------------------------------------- 扰动工具(固定随机种子,结果可复现)
rng = random.Random(7)
def root_alarm_mask(alarms, outages):
"""每条退服的“真根因告警”:同站、清除时间 = 退服恢复时间、类别 = 上报原因、本身不是衍生告警。"""
key = set(zip(outages['ne_id'], outages['end_time'], outages['cause_category']))
parent = alarms['parent_alarm_id'].fillna('')
return pd.Series([(n, c, cat) in key and p == ''
for n, c, cat, p in zip(alarms['ne_id'], alarms['clear_time'], alarms['category'], parent)],
index=alarms.index)
def no_parent(alarms, keep=0.0):
"""去掉父告警关系(keep 为保留比例)。"""
alarms = alarms.copy()
alarms['parent_alarm_id'] = [p if rng.random() < keep else '' for p in alarms['parent_alarm_id'].fillna('')]
return alarms
def drop_root(alarms, outages, fraction):
"""随机删掉一部分真根因告警(网管漏报 / 动环未接入)。"""
mask = root_alarm_mask(alarms, outages)
drop = [i for i in alarms.index[mask] if rng.random() < fraction]
return alarms.drop(drop)
def skew_cell_alarm(alarms, max_seconds=120):
"""“小区退服”告警时刻比工单开始时间晚 5~max_seconds 秒(两套系统时钟不同步)。"""
alarms = alarms.copy()
is_cell = alarms['category'] == '小区退服'
times = pd.to_datetime(alarms.loc[is_cell, 'start_time'])
times = times + pd.to_timedelta([rng.randint(5, max_seconds) for _ in range(len(times))], unit='s')
alarms.loc[is_cell, 'start_time'] = times.dt.strftime(core.TIME_FMT)
return alarms
def jitter_clear(alarms, outages, max_minutes=10):
"""真根因告警的清除时刻与业务恢复不再完全相同,随机偏移 ±max_minutes 分钟。"""
alarms = alarms.copy()
mask = root_alarm_mask(alarms, outages)
times = pd.to_datetime(alarms.loc[mask, 'clear_time'])
times = times + pd.to_timedelta([rng.randint(-max_minutes * 60, max_minutes * 60) for _ in range(len(times))],
unit='s')
alarms.loc[mask, 'clear_time'] = times.dt.strftime(core.TIME_FMT)
return alarms
def add_noise(alarms, outages, fraction):
"""一部分退服开始前 0~12 分钟出现一条无关的短时告警(与 data_gen 的干扰告警同分布)。"""
rows = []
for ne_id, start in zip(outages['ne_id'], pd.to_datetime(outages['start_time'])):
if rng.random() < fraction:
category = rng.choice(['传输中断', '设备硬件', '天馈故障', '机房环境'])
begin = start - pd.Timedelta(minutes=rng.randint(0, 12))
end = begin + pd.Timedelta(minutes=rng.randint(2, 30))
rows.append(dict(alarm_id=f'NZ{len(rows):07d}', ne_id=ne_id, alarm_code='', alarm_name='干扰' + category,
alarm_level='次要', category=category, start_time=begin.strftime(core.TIME_FMT),
clear_time=end.strftime(core.TIME_FMT), duration_s='', cleared='是',
parent_alarm_id=''))
return pd.concat([alarms, pd.DataFrame(rows)], ignore_index=True)
# 比 data_gen 默认更宽的告警提前量(传输/设备也可能提前半小时,停电后备可达 7 小时)
WIDE_LEAD = {'电源故障': (10, 420), '机房环境': (10, 240), '割接施工': (0, 45), '天馈故障': (0, 45),
'传输中断': (0, 30), '设备硬件': (0, 30)}
+115
View File
@@ -0,0 +1,115 @@
"""接口与业务流程测试:登录与权限、导入校验、闭环时限、站型阈值、治理成效、边界情况。
运行:python tests/test_api.py
"""
import io
import pandas as pd
from harness import check, core, finish, load, sample_data # 先导入 harness:它把数据库指向 tests/_work
import app as webapp # noqa: E402
load('api', *sample_data())
core.execute("INSERT INTO users(username, password, role, created_at) VALUES('user', ?, 'user', '')",
(webapp.generate_password_hash('user123'),), changes_data=False)
admin = webapp.app.test_client()
user = webapp.app.test_client()
print('== 1. 登录与权限')
check('错误密码登录失败', admin.post('/login', json={'username': 'admin', 'password': 'x'}).status_code == 400)
check('管理员登录', admin.post('/login', json={'username': 'admin', 'password': 'admin123'}).status_code == 200)
check('普通用户登录', user.post('/login', json={'username': 'user', 'password': 'user123'}).status_code == 200)
check('未登录访问接口返回 401', webapp.app.test_client().get('/api/dashboard').status_code == 401)
for method, url in [('get', '/api/logs'), ('get', '/api/users'), ('post', '/api/analyze'),
('post', '/api/import/preview'), ('post', '/api/outages/batch'), ('put', '/api/site_types/宏站')]:
code = getattr(user, method)(url, json={}).status_code
check(f'普通用户 {method.upper()} {url} 被拒绝', code == 403, f'({code})')
check('普通用户可看看板', user.get('/api/dashboard').status_code == 200)
print('\n== 2. 导入模板校验(模板含 5 行故意写错的数据)')
template = admin.get('/api/template')
preview = admin.post('/api/import/preview', data={'file': (io.BytesIO(template.data), 'template.xlsx')},
content_type='multipart/form-data').get_json()
reasons = {e['row']: e['reason'] for e in preview['errors']}
check('合格 3 行、问题 5 行', (preview['ok'], preview['bad']) == (3, 5), f'({preview["ok"]} / {preview["bad"]})')
expected = {4: '恢复时间早于开始时间', 5: '基站编号缺失', 6: '恢复时间缺失或格式错误', 7: '重复记录', 8: '基站不在站点库中'}
for row, phrase in expected.items():
check(f'第 {row} 行:{phrase}', phrase in reasons.get(row, ''), f'({reasons.get(row)})')
check('上报时长差错 1 条(第 9 行 45 分 vs 实际 405 分)', preview['diff'] == 1)
before = core.query('SELECT COUNT(*) AS n FROM outages')[0]['n']
result = admin.post('/api/import/confirm', json={'token': preview['token']}).get_json()
after = core.query('SELECT COUNT(*) AS n FROM outages')[0]['n']
check('确认后入库 3 条', result['n'] == 3 and after - before == 3)
check('预览令牌只能用一次', admin.post('/api/import/confirm', json={'token': preview['token']}).status_code == 400)
print('\n== 3. 闭环流程与响应时限')
site = core.query('SELECT ne_id FROM sites LIMIT 1')[0]['ne_id']
new = admin.post('/api/outages', json={'ne_id': site, 'start_time': '2026-10-20 08:00', 'end_time': '2026-10-20 08:40',
'status': '待处理'}).get_json()
oid = new['id']
def row():
core.data_changed()
return core.full_frame().set_index('id').loc[oid]
r = row()
check('新增待处理:无接单 / 闭环时间', pd.isna(r['handled_at']) and pd.isna(r['closed_at']))
check('早已超过时限 → 超时未响应', r['sla'] == '超时未响应', f'({r["sla"]},时限 {r["limit_h"]} h)')
check('非法状态被拒绝', user.put(f'/api/outages/{oid}', json={'status': '随便'}).status_code == 400)
check('不存在的记录返回 404', user.put('/api/outages/99999999', json={'status': '处理中'}).status_code == 404)
user.put(f'/api/outages/{oid}', json={'status': '处理中', 'handler': '张工', 'remark': '已派单'})
r = row()
check('转处理中:自动记录接单时间', pd.notna(r['handled_at']) and pd.isna(r['closed_at']))
check('接单晚于时限 → 超时', r['sla'] == '超时', f'({r["sla"]})')
handled_first = r['handled_at']
user.put(f'/api/outages/{oid}', json={'status': '已闭环'})
r = row()
check('转已闭环:自动记录闭环时间,接单时间不变', pd.notna(r['closed_at']) and r['handled_at'] == handled_first)
user.put(f'/api/outages/{oid}', json={'status': '处理中'})
check('重新打开:清空闭环时间', pd.isna(row()['closed_at']))
ids = [x['id'] for x in core.query("SELECT id FROM outages WHERE status = '待处理' LIMIT 3")]
admin.post('/api/outages/batch', json={'ids': ids, 'action': 'status', 'status': '已闭环'})
rows = core.query(f'SELECT handled_at, closed_at FROM outages WHERE id IN ({",".join(map(str, ids))})')
check('批量闭环:都有接单与闭环时间', all(x['handled_at'] and x['closed_at'] for x in rows))
overdue = admin.get('/api/outages', query_string={'overdue': 1, 'size': 5}).get_json()
check('“仅超时未闭环”筛选只返回未闭环且超时的记录',
overdue['total'] > 0 and all(x['status'] != '已闭环' and x['sla'] in ('超时', '超时未响应') for x in overdue['rows']))
k = admin.get('/api/dashboard').get_json()['kpi']
check('看板含及时响应率 / 平均闭环时长 / 超时未闭环',
all(key in k for key in ('response_rate', 'avg_close_h', 'overdue_open')), f'({k["response_rate"]}%,'
f'{k["avg_close_h"]} h,{k["overdue_open"]} 条)')
print('\n== 4. 站型阈值系数')
core.data_changed()
long_before = int(core.full_frame().query("site_type == '室分'")['is_long'].sum())
check('系数越界被拒绝', admin.put('/api/site_types/室分', json={'long_factor': 0, 'freq_factor': 1}).status_code == 400)
admin.put('/api/site_types/室分', json={'long_factor': 0.5, 'freq_factor': 1})
core.data_changed()
frame = core.full_frame()
indoor = frame[frame['site_type'] == '室分']
check('室分系数 0.5:实际阈值 = 场景阈值 × 0.5', (indoor['long_min'] == indoor['scene_long_min'] * 0.5).all())
check('阈值从严后室分超长次数增加', int(indoor['is_long'].sum()) > long_before,
f'({long_before} → {int(indoor["is_long"].sum())})')
macro = frame[frame['site_type'] == '宏站']
check('宏站不受影响', (macro['long_min'] == macro['scene_long_min']).all())
print('\n== 5. 治理成效评估')
effect = admin.get('/api/effect', query_string={'split': '2026-06-01'}).get_json()
check('返回 8 项指标对比', len(effect['rows']) == 8)
check('问题站点改善 + 持平 + 恶化 = 问题站点数',
effect['improved'] + effect['flat'] + effect['worsened'] + sum(s['verdict'] == '—' for s in effect['sites'])
== len(effect['sites']), f'({len(effect["sites"])} 个)')
check('分界日期超出周期返回 400', admin.get('/api/effect', query_string={'split': '2030-01-01'}).status_code == 400)
check('默认分界日期可用', admin.get('/api/effect').status_code == 200)
print('\n== 6. 边界情况')
for url, params in [('/api/dashboard', {'month': '2030-01'}), ('/api/stats', {'month': '2030-01'}),
('/api/association', {'month': '2030-01'}), ('/api/sites/rank', {'site': '不存在的站'}),
('/api/export/stats', {'month': '2030-01'}), ('/api/outages', {'site': '不存在的站'})]:
code = admin.get(url, query_string=params).status_code
check(f'无数据筛选 {url}', code == 200, f'({code})')
finish()
+87
View File
@@ -0,0 +1,87 @@
"""考核指标口径、影响评估与关联量化检验(仓库样例数据)。
运行:python tests/test_kpi.py
"""
import calendar
from harness import check, core, finish, load, sample_data # 先导入 harness:它把数据库指向 tests/_work
import app as webapp # noqa: E402
s, a, o = sample_data()
full = load('kpi', s, a, o)
N = core.site_count()
DAYS = 273 # 样例数据 2026-01-01 ~ 2026-09-30
def calendar_rate(df, n_sites, days):
return round(df['duration_min'].sum() / (n_sites * days * 1440) * 100, 4)
print('== 1. 统计周期')
minutes, months = core.period()
check('全量周期 = 273 天、9 个月', (round(minutes / 1440), months) == (DAYS, 9), f'({minutes / 1440} 天, {months} 月)')
check('全量退服率按日历周期', core.kpis(full, N)['outage_rate'] == calendar_rate(full, N, DAYS))
for m in ['2026-02', '2026-07']:
f = {'month': m}
days = calendar.monthrange(*map(int, m.split('-')))[1]
k = core.kpis(core.frame(f), N, *core.period(f))
check(f'{m} 月退服率按 {days} 天', k['outage_rate'] == calendar_rate(core.frame(f), N, days), f'({k["outage_rate"]}%)')
print('\n== 2. 稀疏筛选')
counts = full.groupby('ne_id').size().sort_values()
ne = counts.index[0]
f = {'site': ne}
df = core.frame(f)
k = core.kpis(df, core.site_count(f), *core.period(f))
check('单站筛选分母为 1 个站', k['sites'] == 1, f'({k["sites"]})')
check('单站退服率按 273 天', k['outage_rate'] == calendar_rate(df, 1, DAYS), f'({k["outage_rate"]}%)')
day = df['start_time'].iloc[0][:10]
f = {'site': ne, 'date_from': day, 'date_to': day}
df = core.frame(f)
k = core.kpis(df, core.site_count(f), *core.period(f))
check('单站单日退服率按 1 站 × 1 天', k['outage_rate'] == calendar_rate(df, 1, 1), f'({k["outage_rate"]}%)')
print('\n== 3. 月均口径:单独查看一个站与全量查看判定一致')
whole = core.site_stats(full, months).set_index('ne_id')
flips = 0
for ne in whole.index:
f = {'site': ne}
one = core.site_stats(core.frame(f), core.period(f)[1]).set_index('ne_id').loc[ne]
flips += (bool(one['long']) != bool(whole.loc[ne, 'long'])) + (bool(one['repeat']) != bool(whole.loc[ne, 'repeat']))
check('超长 / 高频判定翻转数为 0', flips == 0, f'({flips})')
check('重复退服率分母为全部站点', core.kpis(full, N)['repeat_rate'] == round(whole['repeat'].sum() / N * 100, 2))
print('\n== 4. 导出与看板同一口径')
client = webapp.app.test_client()
client.post('/login', json={'username': 'admin', 'password': 'admin123'})
for scene in core.SCENES:
dash = client.get('/api/dashboard', query_string={'scene': scene}).get_json()['kpi']['outage_rate']
f = {'scene': scene}
exported = core.kpis(core.frame(f), core.site_count(f), *core.period(f))['outage_rate']
check(f'{scene} 看板 = 导出', dash == exported, f'({dash}% / {exported}%)')
response = client.get('/api/export/stats', query_string={'scene': '商圈'})
check('导出接口可用', response.status_code == 200)
print('\n== 5. 影响评估')
corr = full['user_hours'].corr(full['duration_min'])
check('受影响用户·小时与时长正相关(修复前受影响用户与时长 r = 0.03)', corr > 0.5, f'(r = {corr:.3f})')
expected = full['affected_users'] * full['duration_min'] / 60
relative = ((full['user_hours'] - expected).abs() / expected.clip(lower=1)).max()
check('用户·小时 = 受影响用户 × 时长', relative < 0.02, f'(最大相对误差 {relative:.4f},来自用户数取整)')
check('覆盖补偿系数在 [0.4, 1]', full['cover_factor'].between(core.COVER_FLOOR, 1).all(),
f'(均值 {full["cover_factor"].mean():.2f})')
isolated = core.site_cover()
check('有邻站的站补偿系数 < 1', min(isolated.values()) < 1)
print(f' 用户·小时合计 {core.kpis(full, N)["user_h"]:,.0f},覆盖补偿系数分布 '
f'{full.drop_duplicates("ne_id")["cover_factor"].value_counts().sort_index().to_dict()}')
print('\n== 6. 告警—退服关联量化')
rows = core.association()
for r in rows:
print(f' {r["category"]:<6} 窗口 {r["window_min"]:>6} 分 告警 {r["alarms"]:>5} 支持度 {r["support"]:>6}% '
f'置信度 {r["confidence"]:>6}% 提升度 {r["lift"]:>6}')
check('6 类告警均有关联指标', len(rows) == 6 and all(r['alarms'] > 0 for r in rows))
check('各类提升度 > 1(告警出现后退服概率高于随机)', all(r['lift'] > 1 for r in rows))
check('关联接口可用', client.get('/api/association').status_code == 200)
finish()
+54
View File
@@ -0,0 +1,54 @@
"""根因判定准确率检验:基线 + 6 组扰动数据集,时间窗只用 1~6 月流水标定,在 7~9 月(留出集)上计分。
运行:python tests/test_rootcause.py
"""
from harness import (WIDE_LEAD, add_noise, check, core, drop_root, finish, generate, jitter_clear, load, no_parent,
sample_data, score, skew_cell_alarm)
SPLIT = '2026-07-01' # 之前的流水用于标定时间窗,之后的用于计分
s, a, o = sample_data()
sw, aw, ow = generate('wide', lead=WIDE_LEAD)
# (名称, 数据, 准确率下限);下限为 None 的只报告不判定——根因告警缺失时,缺失的那部分本来就无法从告警判出
SCENARIOS = [
('S0 基线(仓库样例数据)', (s, a, o), 90),
('S1 去掉父告警(只靠时间窗)', (s, no_parent(a), o), 90),
('S2 小区退服告警时刻偏 5-120 秒', (s, skew_cell_alarm(a), o), 90),
('S3 S1 + 20% 根因告警缺失', (s, drop_root(no_parent(a), o, .2), o), None),
('S4 S1 + 30% 退服前干扰告警', (s, add_noise(no_parent(a), o, .3), o), 90),
('S5 S1 + 告警提前量变宽', (sw, no_parent(aw), ow), 90),
('S6 综合(S5 + 时刻偏移 + 30% 父告警 + 清除时刻 ±10 分 + 15% 缺失 + 25% 干扰)',
# 先删根因告警再扰动清除时刻:drop_root 靠“清除时间 = 恢复时间”识别根因告警
(sw, add_noise(jitter_clear(skew_cell_alarm(drop_root(no_parent(aw, keep=.3), ow, .15)), ow), ow, .25), ow),
None),
]
summary = []
for name, (sites, alarms, outages), floor in SCENARIOS:
df = load('rootcause', sites, alarms, outages, calibrate_until=SPLIT)
test = df[df['start_time'] >= SPLIT]
accuracy, rows = score(test)
other = (test['root_cause'] == '其他').mean() * 100
print(f'\n== {name}\n 留出集 {len(test)} 条,准确率 {accuracy}%,判为“其他” {other:.1f}%')
print(' 类别 precision recall F1')
for cause, precision, recall, f1 in rows:
print(f' {cause:<8} {precision:>7} {recall:>7} {f1:>6}')
confidence = test.groupby('confidence').apply(lambda x: (x['cause_category'] == x['root_cause']).mean() * 100,
include_groups=False).round(1)
print(' 各置信度准确率:', confidence.to_dict(), ' 条数:', test['confidence'].value_counts().to_dict())
print(' 时间窗:', {w['category']: w['window_min'] for w in core.window_table()})
summary.append((name, accuracy))
if floor is not None:
check(f'{name.split()[0]} 准确率 ≥ {floor}%', accuracy >= floor, f'({accuracy}%)')
else:
# 缺失场景的理论上限:真根因告警仍在的比例;判定应接近它,且“高/中”置信度的判定仍应可靠
reliable = test[test['confidence'].isin(['高', '中'])]
rel_acc = (reliable['cause_category'] == reliable['root_cause']).mean() * 100
check(f'{name.split()[0]} 高/中置信度判定准确率 ≥ 90%', rel_acc >= 90, f'({rel_acc:.1f}%,'
f'占 {len(reliable) / len(test) * 100:.0f}%)')
print('\n汇总(留出集准确率):')
for name, accuracy in summary:
print(f' {accuracy:>6}% {name}')
finish()