热点聚类.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "import re\n",
    "import time\n",
    "import jieba\n",
    "import jieba.posseg as pseg\n",
    "import json\n",
    "from datetime import datetime\n",
    "from datetime import timedelta\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>time</th>\n",
       "      <th>content</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>A市西湖建筑集团占道施工有安全隐患</td>\n",
       "      <td>2020/1/6 12:09:38</td>\n",
       "      <td>\\n\\t\\t\\t\\t\\t\\n\\t\\t\\t\\t\\tA3区大道西行便道，未管所路口至加油站路段，...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>A市在水一方大厦人为烂尾多年，安全隐患严重</td>\n",
       "      <td>2020/1/4 11:17:46</td>\n",
       "      <td>\\n\\t\\t\\t\\t\\t\\n\\t\\t\\t\\t\\t位于书院路主干道的在水一方大厦一楼至四楼人为...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>A3区杜鹃文苑小区外的非法汽车检测站要开业了！</td>\n",
       "      <td>2018/12/26 10:13:37</td>\n",
       "      <td>\\n\\t\\t\\t\\t\\t\\n\\t\\t\\t\\t\\tA市政府、市交警支队、市安监局、市环保局、A...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>民工在A6区明发国际工地受伤，工地方拒绝支付医疗费</td>\n",
       "      <td>2018/7/17 7:13:58</td>\n",
       "      <td>\\n\\t\\t\\t\\t\\t\\n\\t\\t\\t\\t\\t胡书记，您好，感谢您百忙之中查看这份留言。我...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>K8县丁字街的商户乱摆摊</td>\n",
       "      <td>2017/10/14 14:25:14</td>\n",
       "      <td>\\n\\t\\t\\t\\t\\t\\n\\t\\t\\t\\t\\t　　K8县丁字街的商户乱摆摊，前段时间丁字街...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                       title                 time  \\\n",
       "0          A市西湖建筑集团占道施工有安全隐患    2020/1/6 12:09:38   \n",
       "1      A市在水一方大厦人为烂尾多年，安全隐患严重    2020/1/4 11:17:46   \n",
       "2    A3区杜鹃文苑小区外的非法汽车检测站要开业了！  2018/12/26 10:13:37   \n",
       "3  民工在A6区明发国际工地受伤，工地方拒绝支付医疗费    2018/7/17 7:13:58   \n",
       "4               K8县丁字街的商户乱摆摊  2017/10/14 14:25:14   \n",
       "\n",
       "                                             content  \n",
       "0  \\n\\t\\t\\t\\t\\t\\n\\t\\t\\t\\t\\tA3区大道西行便道，未管所路口至加油站路段，...  \n",
       "1  \\n\\t\\t\\t\\t\\t\\n\\t\\t\\t\\t\\t位于书院路主干道的在水一方大厦一楼至四楼人为...  \n",
       "2  \\n\\t\\t\\t\\t\\t\\n\\t\\t\\t\\t\\tA市政府、市交警支队、市安监局、市环保局、A...  \n",
       "3  \\n\\t\\t\\t\\t\\t\\n\\t\\t\\t\\t\\t胡书记，您好，感谢您百忙之中查看这份留言。我...  \n",
       "4  \\n\\t\\t\\t\\t\\t\\n\\t\\t\\t\\t\\t　　K8县丁字街的商户乱摆摊，前段时间丁字街...  "
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "filepath =  'data/data.csv'\n",
    "#filepath =  'data/news/news_df.csv'\n",
    "df = pd.read_csv(filepath,index_col=0)\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def data_filter(df):\n",
    "    \"\"\"数据过滤\"\"\"\n",
    "    # 过滤掉没有内容的新闻\n",
    "    df = df[df['content'] != ''].copy()\n",
    "    df = df.dropna(subset=['content']).copy()\n",
    "    # 去重\n",
    "    #df = df.drop_duplicates(subset=['title'])\n",
    "    df = df.reset_index(drop=True)\n",
    "    return df\n",
    "\n",
    "\n",
    "def get_data(df, last_time, delta):\n",
    "    \"\"\"\n",
    "    获取某段时间的新闻数据\n",
    "    :param df: 原始数据\n",
    "    :param last_time: 指定要获取数据的最后时间\n",
    "    :param delta: 时间间隔\n",
    "    :return: last_time前timedelta的数据\n",
    "    \"\"\"\n",
    "    last_time = datetime.strptime(last_time, '%Y/%m/%d %H:%M:%S')\n",
    "    delta = timedelta(delta)\n",
    "    try:\n",
    "        df['time'] = df['time'].map(lambda x: datetime.strptime(x, '%Y/%m/%d %H:%M:%S'))\n",
    "    except TypeError:\n",
    "        pass\n",
    "    df = df[df['time'].map(lambda x: (x <= last_time) and (x > last_time - delta))].copy()\n",
    "    print('df.shape=', df.shape)\n",
    "    if df.shape[0] == 0:\n",
    "        print('No Data!')\n",
    "        return df\n",
    "    df = df.sort_values(by=['time'], ascending=[0])\n",
    "    df['time'] = df['time'].map(lambda x: datetime.strftime(x, '%Y/%m/%d %H:%M:%S'))\n",
    "    df = df.reset_index(drop=True)\n",
    "    return df\n",
    "\n",
    "\n",
    "def clean_title_blank(title):\n",
    "    \"\"\"清理新闻标题空白\"\"\"\n",
    "    # 英文大写转小写\n",
    "    title = title.upper()\n",
    "    # 清理未知字符\n",
    "    title = re.sub(r'\\?+', ' ', title)\n",
    "    # 清理空白字符\n",
    "    title = re.sub(r'\\u3000', '', title)\n",
    "    title = title.strip()\n",
    "    title = re.sub(r'\\s+', ' ', title)\n",
    "    title = re.sub(r'([|：])+ ', r'\\1', title)\n",
    "    title = re.sub(r' ([|：])+', r'\\1', title)\n",
    "    return title\n",
    "\n",
    "\n",
    "def clean_content_blank(content):\n",
    "    \"\"\"清理新闻内容空白\"\"\"\n",
    "    # 清理未知字符\n",
    "    content = re.sub(r'\\?+', ' ', content)\n",
    "    # 清理空白字符\n",
    "    content = re.sub(r'\\u3000', '', content)\n",
    "    content = content.strip()\n",
    "    content = re.sub(r'[ \\t\\r\\f]+', ' ', content)\n",
    "    content = re.sub(r'\\n ', '\\n', content)\n",
    "    content = re.sub(r' \\n', '\\n', content)\n",
    "    content = re.sub(r'\\n+', '\\n', content)\n",
    "    return content\n",
    "\n",
    "\n",
    "def clean_content(content):\n",
    "    \"\"\"清理新闻内容\"\"\"\n",
    "    # 清理新闻内容空白\n",
    "    content = clean_content_blank(content)\n",
    "    # 英文大写转小写\n",
    "    content = content.upper()\n",
    "    # 清理超链接\n",
    "    content = re.sub(r'https?://[-A-Za-z0-9+&@#/%?=~_|!:,.;]+[-A-Za-z0-9+&@#/%=~_|]', '', content)\n",
    "    # 清理来源等和内容无关的文字\n",
    "    texts = []\n",
    "    for text in texts:\n",
    "        content = re.sub(text, '', content)\n",
    "    content = re.sub(r'\\n+', '\\n', content)\n",
    "    return content\n",
    "\n",
    "\n",
    "def get_num_en_ch(text):\n",
    "    \"\"\"提取数字英文中文\"\"\"\n",
    "    text = re.sub(r'[^0-9A-Za-z\\u4E00-\\u9FFF]+', ' ', text)\n",
    "    text = text.strip()\n",
    "    return text\n",
    "\n",
    "\n",
    "def pseg_cut(text, userdict_path=None):\n",
    "    \"\"\"\n",
    "    词性标注\n",
    "    :param text: string，原文本数据\n",
    "    :param userdict_path: string，用户词词典路径，默认为None\n",
    "    :return: list， 分词后词性标注的列表\n",
    "    \"\"\"\n",
    "    if userdict_path is not None:\n",
    "        jieba.load_userdict(userdict_path)\n",
    "    words = pseg.lcut(text)\n",
    "    return words\n",
    "\n",
    "\n",
    "def get_words_by_flags(words, flags=None):\n",
    "    \"\"\"\n",
    "    获取指定词性的词\n",
    "    :param words: list， 分词后词性标注的列表\n",
    "    :param flags: list， 词性标注，默认为提取名词和动词\n",
    "    :return: list， 指定词性的词\n",
    "    \"\"\"\n",
    "    flags = ['n.*', 'v.*'] if flags is None else flags\n",
    "    words = [w for w, f in words if w != ' ' and re.match('|'.join(['(%s$)' % flag for flag in flags]), f)]\n",
    "    return words\n",
    "\n",
    "\n",
    "def userdict_cut(text, userdict_path=None):\n",
    "    \"\"\"\n",
    "    对文本进行jieba分词\n",
    "    如果使用用户词词典，那么使用用户词词典进行jieba分词\n",
    "    \"\"\"\n",
    "    if userdict_path is not None:\n",
    "        jieba.load_userdict(userdict_path)\n",
    "    words = jieba.cut(text)\n",
    "    return words\n",
    "\n",
    "\n",
    "def stop_words_cut(words, stop_words_path):\n",
    "    \"\"\"停用词处理\"\"\"\n",
    "    with open(stop_words_path, 'r', encoding='utf-8') as f:\n",
    "        stopwords = [line.strip() for line in f.readlines()]\n",
    "        stopwords.append(' ')\n",
    "        words = [word for word in words if word not in stopwords]\n",
    "    return words\n",
    "\n",
    "\n",
    "def disambiguation_cut(words, disambiguation_dict_path):\n",
    "    \"\"\"消歧词典\"\"\"\n",
    "    with open(disambiguation_dict_path, 'r', encoding='utf-8') as f:\n",
    "        disambiguation_dict = json.load(f)\n",
    "        words = [(disambiguation_dict[word]\n",
    "                  if disambiguation_dict.get(word) else word) for word in words]\n",
    "    return words\n",
    "\n",
    "\n",
    "def individual_character_cut(words, individual_character_dict_path):\n",
    "    \"\"\"删除无用单字\"\"\"\n",
    "    with open(individual_character_dict_path, 'r', encoding='utf-8') as f:\n",
    "        individual_character = [line.strip() for line in f.readlines()]\n",
    "        words = [word for word in words\n",
    "                 if ((len(word) > 1) or ((len(word) == 1) and (word in individual_character)))]\n",
    "    return words\n",
    "\n",
    "\n",
    "def document2txt(raw_document, userdict_path, text_path):\n",
    "    \"\"\"文本分词并保存为txt文件\"\"\"\n",
    "    document = clean_content_blank(raw_document)\n",
    "    document = document.upper()\n",
    "    document_cut = userdict_cut(document, userdict_path)\n",
    "    result = ' '.join(document_cut)\n",
    "    result = re.sub(r' +', ' ', result)\n",
    "    result = re.sub(r' \\n ', '\\n', result)\n",
    "    with open(text_path, 'w+', encoding='utf-8') as f:\n",
    "        f.write(result)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Building prefix dict from the default dictionary ...\n",
      "Loading model from cache C:\\Users\\sun\\AppData\\Local\\Temp\\jieba.cache\n",
      "Loading model cost 0.910 seconds.\n",
      "Prefix dict has been built successfully.\n"
     ]
    }
   ],
   "source": [
    "# 数据清洗\n",
    "\n",
    "# 去除空白\n",
    "df['title_'] = df['title'].map(lambda x: clean_title_blank(x))\n",
    "df['content'] = df['content'].map(lambda x: clean_content(x))\n",
    "df['content_'] = df['content'].map(lambda x: get_num_en_ch(x))\n",
    "df['content_cut'] = df['content_'].map(lambda x: pseg_cut(\n",
    "    x, userdict_path='data/extra_dict/self_userdict.txt'))\n",
    "df['content_cut'] = df['content_cut'].map(lambda x: get_words_by_flags(\n",
    "    x, flags=['n.*', 'v.*', 'eng', 't', 's', 'j', 'l', 'i']))\n",
    "\n",
    "df['content_cut'] = df['content_cut'].map(lambda x: stop_words_cut(\n",
    "    x, 'data/extra_dict/self_stop_words.txt'))\n",
    "\n",
    "df['content_cut'] = df['content_cut'].map(lambda x: disambiguation_cut(\n",
    "    x,'data/extra_dict/self_disambiguation_dict.json'))\n",
    "df['content_cut'] = df['content_cut'].map(lambda x: individual_character_cut(\n",
    "    x,  'data/extra_dict/self_individual_character_dict.txt'))\n",
    "df['content_'] = df['content_cut'].map(lambda x: ' '.join(x))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>time</th>\n",
       "      <th>content</th>\n",
       "      <th>title_</th>\n",
       "      <th>content_</th>\n",
       "      <th>content_cut</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>A市西湖建筑集团占道施工有安全隐患</td>\n",
       "      <td>2020/1/6 12:09:38</td>\n",
       "      <td>A3区大道西行便道，未管所路口至加油站路段，人行道包括路灯杆，被圈西湖建筑集团燕子山安置房项...</td>\n",
       "      <td>A市西湖建筑集团占道施工有安全隐患</td>\n",
       "      <td>大道 西行 路口 加油站 路段 人行道 包括 路灯 西湖 建筑 集团 燕子 安置 项目 施工...</td>\n",
       "      <td>[大道, 西行, 路口, 加油站, 路段, 人行道, 包括, 路灯, 西湖, 建筑, 集团,...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>A市在水一方大厦人为烂尾多年，安全隐患严重</td>\n",
       "      <td>2020/1/4 11:17:46</td>\n",
       "      <td>位于书院路主干道的在水一方大厦一楼至四楼人为拆除水、电等设施后，烂尾多年，用护栏围着，不但占...</td>\n",
       "      <td>A市在水一方大厦人为烂尾多年，安全隐患严重</td>\n",
       "      <td>位于 书院 主干道 在水一方 大厦 一楼 四楼 人为 拆除 设施 烂尾 护栏 围着 占用 人...</td>\n",
       "      <td>[位于, 书院, 主干道, 在水一方, 大厦, 一楼, 四楼, 人为, 拆除, 设施, 烂尾...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   title               time  \\\n",
       "0      A市西湖建筑集团占道施工有安全隐患  2020/1/6 12:09:38   \n",
       "1  A市在水一方大厦人为烂尾多年，安全隐患严重  2020/1/4 11:17:46   \n",
       "\n",
       "                                             content                 title_  \\\n",
       "0  A3区大道西行便道，未管所路口至加油站路段，人行道包括路灯杆，被圈西湖建筑集团燕子山安置房项...      A市西湖建筑集团占道施工有安全隐患   \n",
       "1  位于书院路主干道的在水一方大厦一楼至四楼人为拆除水、电等设施后，烂尾多年，用护栏围着，不但占...  A市在水一方大厦人为烂尾多年，安全隐患严重   \n",
       "\n",
       "                                            content_  \\\n",
       "0  大道 西行 路口 加油站 路段 人行道 包括 路灯 西湖 建筑 集团 燕子 安置 项目 施工...   \n",
       "1  位于 书院 主干道 在水一方 大厦 一楼 四楼 人为 拆除 设施 烂尾 护栏 围着 占用 人...   \n",
       "\n",
       "                                         content_cut  \n",
       "0  [大道, 西行, 路口, 加油站, 路段, 人行道, 包括, 路灯, 西湖, 建筑, 集团,...  \n",
       "1  [位于, 书院, 主干道, 在水一方, 大厦, 一楼, 四楼, 人为, 拆除, 设施, 烂尾...  "
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from collections import Counter\n",
    "\n",
    "def flat(l):\n",
    "    \"\"\"平展多维列表\"\"\"\n",
    "    for k in l:\n",
    "        if not isinstance(k, (list, tuple)):\n",
    "            yield k\n",
    "        else:\n",
    "            yield from flat(k)\n",
    "\n",
    "\n",
    "def get_word_library(list1):\n",
    "    \"\"\"\n",
    "    获得词库\n",
    "    :param list1: 一维或多维词列表\n",
    "    :return: list，所有词去重之后的列表\n",
    "    \"\"\"\n",
    "    list2 = flat(list1)\n",
    "    list3 = list(set(list2))\n",
    "    return list3\n",
    "\n",
    "\n",
    "def get_single_frequency_words(list1):\n",
    "    \"\"\"\n",
    "    获得单频词列表\n",
    "    :param list1: 一维或多维词列表\n",
    "    :return: list，所有只出现一次的词组成的列表\n",
    "    \"\"\"\n",
    "    list2 = flat(list1)\n",
    "    cnt = Counter(list2)\n",
    "    list3 = [i for i in cnt if cnt[i] == 1]\n",
    "    return list3\n",
    "\n",
    "\n",
    "def get_most_common_words(list1, top_n=None, min_frequency=1):\n",
    "    \"\"\"\n",
    "    获取最常见的词组成的列表\n",
    "    :param list1: 一维或多维词列表\n",
    "    :param top_n: 指定最常见的前n个词，默认为None\n",
    "    :param min_frequency: 指定最小频数，默认为1\n",
    "    :return: list，最常见的前n个词组成的列表\n",
    "    \"\"\"\n",
    "    list2 = flat(list1)\n",
    "    cnt = Counter(list2)\n",
    "    list3 = [i[0] for i in cnt.most_common(top_n) if cnt[i[0]] >= min_frequency]\n",
    "    return list3\n",
    "\n",
    "\n",
    "def get_num_of_value_no_repeat(list1):\n",
    "    \"\"\"\n",
    "    获取列表中不重复的值的个数\n",
    "    :param list1: 列表\n",
    "    :return: int，列表中不重复的值的个数\n",
    "    \"\"\"\n",
    "    num = len(set(list1))\n",
    "    return num\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pickle\n",
    "from collections import Counter\n",
    "from sklearn.feature_extraction.text import CountVectorizer, TfidfVectorizer\n",
    "from sklearn.cluster import DBSCAN, KMeans\n",
    "from sklearn.decomposition import PCA\n",
    "from sklearn.manifold import TSNE\n",
    "from wordcloud import WordCloud\n",
    "from textrank4zh import TextRank4Sentence\n",
    "from gensim.models import word2vec\n",
    "\n",
    "\n",
    "def feature_extraction(series, vectorizer='CountVectorizer', vec_args=None):\n",
    "    \"\"\"\n",
    "    对原文本进行特征提取\n",
    "    :param series: pd.Series，原文本\n",
    "    :param vectorizer: string，矢量化器，如'CountVectorizer'或者'TfidfVectorizer'\n",
    "    :param vec_args: dict，矢量化器参数\n",
    "    :return: 稀疏矩阵\n",
    "    \"\"\"\n",
    "    vec_args = {'max_df': 1.0, 'min_df': 1} if vec_args is None else vec_args\n",
    "    vec_args_list = ['%s=%s' % (i[0],\n",
    "                                \"'%s'\" % i[1] if isinstance(i[1], str) else i[1]\n",
    "                                ) for i in vec_args.items()]\n",
    "    vec_args_str = ','.join(vec_args_list)\n",
    "    vectorizer1 = eval(\"%s(%s)\" % (vectorizer, vec_args_str))\n",
    "    matrix = vectorizer1.fit_transform(series)\n",
    "    return matrix\n",
    "\n",
    "\n",
    "def get_cluster(matrix, cluster='DBSCAN', cluster_args=None):\n",
    "    \"\"\"\n",
    "    对数据进行聚类，获取训练好的聚类器\n",
    "    :param matrix: 稀疏矩阵\n",
    "    :param cluster: string，聚类器\n",
    "    :param cluster_args: dict，聚类器参数\n",
    "    :return: 训练好的聚类器\n",
    "    \"\"\"\n",
    "    cluster_args = {'eps': 0.5, 'min_samples': 5, 'metric': 'cosine'} if cluster_args is None else cluster_args\n",
    "    cluster_args_list = ['%s=%s' % (i[0],\n",
    "                                    \"'%s'\" % i[1] if isinstance(i[1], str) else i[1]\n",
    "                                    ) for i in cluster_args.items()]\n",
    "    cluster_args_str = ','.join(cluster_args_list)\n",
    "    cluster1 = eval(\"%s(%s)\" % (cluster, cluster_args_str))\n",
    "    cluster1 = cluster1.fit(matrix)\n",
    "    return cluster1\n",
    "\n",
    "\n",
    "def get_labels(cluster):\n",
    "    \"\"\"\n",
    "    获取聚类标签\n",
    "    :param cluster: 训练好的聚类器\n",
    "    :return: list，聚类标签\n",
    "    \"\"\"\n",
    "    labels = cluster.labels_\n",
    "    return labels\n",
    "\n",
    "\n",
    "def label2rank(labels_list):\n",
    "    \"\"\"\n",
    "    按标签的数量将标签转换为排行\n",
    "    :param labels_list: list，聚类标签\n",
    "    :return: list，聚类排行\n",
    "    \"\"\"\n",
    "    series = pd.Series(labels_list)\n",
    "    list1 = series[series != -1].tolist()\n",
    "    n = len(set(list1))\n",
    "    cnt = Counter(list1)\n",
    "    key = [cnt.most_common()[i][0] for i in range(n)]\n",
    "    value = [i for i in range(1, n + 1)]\n",
    "    my_dict = dict(zip(key, value))\n",
    "    my_dict[-1] = -1\n",
    "    rank_list = [my_dict[i] for i in labels_list]\n",
    "    return rank_list\n",
    "\n",
    "\n",
    "def get_non_outliers_data(df, label_column='label'):\n",
    "    \"\"\"获取属于某个聚类簇的数据\"\"\"\n",
    "    df = df[df[label_column] != -1].copy()\n",
    "    return df\n",
    "\n",
    "\n",
    "def get_data_sort_labelnum(df, label_column='label', top=1):\n",
    "    \"\"\"\n",
    "    获取按标签数量排行的第top组数据\n",
    "    :param df: pd.DataFrame，带有标签列的数据\n",
    "    :param label_column: string，标签列名\n",
    "    :param top: int\n",
    "    :return: pd.DataFrame，前top组的数据\n",
    "    \"\"\"\n",
    "    assert top > 0, 'top不能小于等于0！'\n",
    "    labels = df[label_column].tolist()\n",
    "    cnt = Counter(labels)\n",
    "    label = cnt.most_common()[top - 1][0] if top <= len(set(labels)) else -2\n",
    "    df = df[df[label_column] == label].copy() if label != -2 else pd.DataFrame(columns=df.columns)\n",
    "    return df\n",
    "\n",
    "\n",
    "def list2wordcloud(list1, save_path, font_path):\n",
    "    \"\"\"\n",
    "    将文本做成词云\n",
    "    :param list1: list，文本列表\n",
    "    :param save_path: string，词云图片保存的路径\n",
    "    :param font_path: string，用于制作词云所需的字体路径\n",
    "    \"\"\"\n",
    "    text = ' '.join(list1)\n",
    "    wc = WordCloud(font_path=font_path, width=800, height=600, margin=2,\n",
    "                   ranks_only=True, max_words=200, collocations=False).generate(text)\n",
    "    wc.to_file(save_path)\n",
    "\n",
    "\n",
    "def get_key_sentences(text, num=1):\n",
    "    \"\"\"\n",
    "    利用textrank算法，获取文本摘要\n",
    "    :param text: string，原文本\n",
    "    :param num: int，指定摘要条数\n",
    "    :return: string，文本摘要\n",
    "    \"\"\"\n",
    "    tr4s = TextRank4Sentence(delimiters='\\n')\n",
    "    tr4s.analyze(text=text, source='all_filters')\n",
    "    abstract = '\\n'.join([item.sentence for item in tr4s.get_key_sentences(num=num)])\n",
    "    return abstract\n",
    "\n",
    "\n",
    "def feature_reduction(matrix, pca_n_components=50, tsne_n_components=2):\n",
    "    \"\"\"降维\"\"\"\n",
    "    data_pca = PCA(n_components=pca_n_components).fit_transform(matrix) if pca_n_components is not None else matrix\n",
    "    data_pca_tsne = TSNE(n_components=tsne_n_components).fit_transform(\n",
    "        data_pca) if tsne_n_components is not None else data_pca\n",
    "    print('data_pca_tsne.shape=', data_pca_tsne.shape)\n",
    "    return data_pca_tsne\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "\n",
    "def get_wordvec(model, word):\n",
    "    \"\"\"查询词是否在词库中\"\"\"\n",
    "    try:\n",
    "        model.wv.get_vector(word)\n",
    "        return True\n",
    "    except:\n",
    "        return False\n",
    "\n",
    "\n",
    "def get_word_and_wordvec(model, words):\n",
    "    \"\"\"获取输入词的词和对应的词向量\"\"\"\n",
    "    word_list = [i for i in words if get_wordvec(model, i)]\n",
    "    wordvec_list = [model.wv[i].tolist() for i in words if get_wordvec(model, i)]\n",
    "    return word_list, wordvec_list\n",
    "\n",
    "\n",
    "def get_top_words(words, label, label_num):\n",
    "    \"\"\"获得每个类中的前30个词\"\"\"\n",
    "    df = pd.DataFrame()\n",
    "    df['word'] = words\n",
    "    df['label'] = label\n",
    "    for i in range(label_num):\n",
    "        df_ = df[df['label'] == i]\n",
    "        print(df_['word'][:30])\n",
    "\n",
    "\n",
    "def save_model(model, model_path):\n",
    "    \"\"\"保存模型\"\"\"\n",
    "    with open(model_path, 'wb') as f:\n",
    "        pickle.dump(model, f)\n",
    "\n",
    "\n",
    "def load_model(model_path):\n",
    "    \"\"\"加载模型\"\"\"\n",
    "    with open(model_path, 'rb') as f:\n",
    "        model = pickle.load(f)\n",
    "    return model\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "word_library_list = get_word_library(df['content_cut'])\n",
    "single_frequency_words_list = get_single_frequency_words(df['content_cut'])\n",
    "# 设置最大特征词数\n",
    "max_features = (len(word_library_list) - len(single_frequency_words_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 转换为TDIDF\n",
    "matrix = feature_extraction(df['content_'], vectorizer='TfidfVectorizer',\n",
    "                                         vec_args={'max_df': 0.95, 'min_df': 1, 'max_features': max_features})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 建立模型 最大边距/ 最小个数\n",
    "eps_var = 0.6\n",
    "min_samples_var = 3\n",
    "dbscan = get_cluster(matrix, cluster='DBSCAN',\n",
    "                    cluster_args={'eps': eps_var, 'min_samples': min_samples_var, 'metric': 'cosine'})"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "labels = get_labels(dbscan)\n",
    "df['label'] = labels\n",
    "ranks = label2rank(labels)\n",
    "df['rank'] = ranks"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>time</th>\n",
       "      <th>content</th>\n",
       "      <th>title_</th>\n",
       "      <th>content_</th>\n",
       "      <th>content_cut</th>\n",
       "      <th>label</th>\n",
       "      <th>rank</th>\n",
       "      <th>matrix</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>A市西湖建筑集团占道施工有安全隐患</td>\n",
       "      <td>2020/1/6 12:09:38</td>\n",
       "      <td>A3区大道西行便道，未管所路口至加油站路段，人行道包括路灯杆，被圈西湖建筑集团燕子山安置房项...</td>\n",
       "      <td>A市西湖建筑集团占道施工有安全隐患</td>\n",
       "      <td>大道 西行 路口 加油站 路段 人行道 包括 路灯 西湖 建筑 集团 燕子 安置 项目 施工...</td>\n",
       "      <td>[大道, 西行, 路口, 加油站, 路段, 人行道, 包括, 路灯, 西湖, 建筑, 集团,...</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>A市在水一方大厦人为烂尾多年，安全隐患严重</td>\n",
       "      <td>2020/1/4 11:17:46</td>\n",
       "      <td>位于书院路主干道的在水一方大厦一楼至四楼人为拆除水、电等设施后，烂尾多年，用护栏围着，不但占...</td>\n",
       "      <td>A市在水一方大厦人为烂尾多年，安全隐患严重</td>\n",
       "      <td>位于 书院 主干道 在水一方 大厦 一楼 四楼 人为 拆除 设施 烂尾 护栏 围着 占用 人...</td>\n",
       "      <td>[位于, 书院, 主干道, 在水一方, 大厦, 一楼, 四楼, 人为, 拆除, 设施, 烂尾...</td>\n",
       "      <td>-1</td>\n",
       "      <td>-1</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                   title               time  \\\n",
       "0      A市西湖建筑集团占道施工有安全隐患  2020/1/6 12:09:38   \n",
       "1  A市在水一方大厦人为烂尾多年，安全隐患严重  2020/1/4 11:17:46   \n",
       "\n",
       "                                             content                 title_  \\\n",
       "0  A3区大道西行便道，未管所路口至加油站路段，人行道包括路灯杆，被圈西湖建筑集团燕子山安置房项...      A市西湖建筑集团占道施工有安全隐患   \n",
       "1  位于书院路主干道的在水一方大厦一楼至四楼人为拆除水、电等设施后，烂尾多年，用护栏围着，不但占...  A市在水一方大厦人为烂尾多年，安全隐患严重   \n",
       "\n",
       "                                            content_  \\\n",
       "0  大道 西行 路口 加油站 路段 人行道 包括 路灯 西湖 建筑 集团 燕子 安置 项目 施工...   \n",
       "1  位于 书院 主干道 在水一方 大厦 一楼 四楼 人为 拆除 设施 烂尾 护栏 围着 占用 人...   \n",
       "\n",
       "                                         content_cut  label  rank  \\\n",
       "0  [大道, 西行, 路口, 加油站, 路段, 人行道, 包括, 路灯, 西湖, 建筑, 集团,...     -1    -1   \n",
       "1  [位于, 书院, 主干道, 在水一方, 大厦, 一楼, 四楼, 人为, 拆除, 设施, 烂尾...     -1    -1   \n",
       "\n",
       "                                              matrix  \n",
       "0  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...  \n",
       "1  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...  "
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "共有 5 类\n"
     ]
    }
   ],
   "source": [
    "df['matrix'] = matrix.toarray().tolist()\n",
    "\n",
    "# 筛选出所有聚类的点\n",
    "df_non_outliers = df[df['label'] != -1].copy()\n",
    "\n",
    "# 类别数量\n",
    "rank_num = get_num_of_value_no_repeat(df_non_outliers['rank'])\n",
    "print(\"共有\",rank_num,\"类\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>title</th>\n",
       "      <th>time</th>\n",
       "      <th>content</th>\n",
       "      <th>title_</th>\n",
       "      <th>content_</th>\n",
       "      <th>content_cut</th>\n",
       "      <th>label</th>\n",
       "      <th>rank</th>\n",
       "      <th>matrix</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>134</td>\n",
       "      <td>请降低出租车行业门槛</td>\n",
       "      <td>2014/7/18 8:31:37</td>\n",
       "      <td>出租车司机收入水平下降，黑车越来越多。不利于人民富足，而且增加政府行政成本。解决问题的关键在...</td>\n",
       "      <td>请降低出租车行业门槛</td>\n",
       "      <td>出租车 司机 收入水平 下降 黑车 不利于 人民富足 增加 政府 行政 成本 解决问题 关键...</td>\n",
       "      <td>[出租车, 司机, 收入水平, 下降, 黑车, 不利于, 人民富足, 增加, 政府, 行政,...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>136</td>\n",
       "      <td>对A市出租车管理的建议</td>\n",
       "      <td>2011/11/18 15:48:58</td>\n",
       "      <td>作为一个省级政府机关的公务员，近二十年来走遍中国的大城市，只有A市的出租车没有一点规矩，最不...</td>\n",
       "      <td>对A市出租车管理的建议</td>\n",
       "      <td>政府 机关 公务员 走遍 中国 大城市 出租车 规矩 文明 出游 乘兴 出租车 机场 火车站...</td>\n",
       "      <td>[政府, 机关, 公务员, 走遍, 中国, 大城市, 出租车, 规矩, 文明, 出游, 乘兴...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>138</td>\n",
       "      <td>我建议A市出租车最好搞个体经营</td>\n",
       "      <td>2013/3/23 20:52:51</td>\n",
       "      <td>李总理讲：凡是市场化能解决的，都要由市场决定，政府不要插手（大意）。又讲：触动利益比触动灵魂...</td>\n",
       "      <td>我建议A市出租车最好搞个体经营</td>\n",
       "      <td>总理 市场化 解决 市场 政府 插手 触动 利益 触动 灵魂 出租车 个体经营 驾驶员 交警...</td>\n",
       "      <td>[总理, 市场化, 解决, 市场, 政府, 插手, 触动, 利益, 触动, 灵魂, 出租车,...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>140</td>\n",
       "      <td>请书记为M4市出租车司机做主！</td>\n",
       "      <td>2014/3/2 20:09:32</td>\n",
       "      <td>我是一名普通的出租车司机，出生在一个普通的工人家中，一家五口就靠父亲那点微薄的工资来维持生活...</td>\n",
       "      <td>请书记为M4市出租车司机做主！</td>\n",
       "      <td>出租车 司机 出生 工人 家中 父亲 工资 生活 家中 无田无土 借钱 出租车 不到 时间 ...</td>\n",
       "      <td>[出租车, 司机, 出生, 工人, 家中, 父亲, 工资, 生活, 家中, 无田无土, 借钱...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>141</td>\n",
       "      <td>请问G市网约车政策何时出台和落实</td>\n",
       "      <td>2017/7/12 11:39:24</td>\n",
       "      <td>我是一名出租车司机，从事出租车行业已经有八年的时间了，由于网约车的兴起，现在出租车的收入大不...</td>\n",
       "      <td>请问G市网约车政策何时出台和落实</td>\n",
       "      <td>出租车 司机 出租车 行业 八年 时间 网约车 兴起 出租车 收入 大不如前 收入 保安 工...</td>\n",
       "      <td>[出租车, 司机, 出租车, 行业, 八年, 时间, 网约车, 兴起, 出租车, 收入, 大...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>142</td>\n",
       "      <td>网约车无序，A市出租车司机的权益谁来保障？</td>\n",
       "      <td>2019/7/15 20:11:28</td>\n",
       "      <td>黑车猖狂，网约车无序，的士夹缝中求生存，A市一有大事就狠抓合规出租车，可根本问题却是出租车司...</td>\n",
       "      <td>网约车无序，A市出租车司机的权益谁来保障？</td>\n",
       "      <td>黑车 网约 无序 的士 夹缝 求生存 狠抓 合规 出租车 出租车 司机 权益 保障</td>\n",
       "      <td>[黑车, 网约, 无序, 的士, 夹缝, 求生存, 狠抓, 合规, 出租车, 出租车, 司机...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>144</td>\n",
       "      <td>B9市出租车（的士）随口说价，不打表和提供正规发票</td>\n",
       "      <td>2015/6/28 9:26:35</td>\n",
       "      <td>B9市客运管理处：出租车是城市文明的窗口，市民交通出行最基本的保障，现针对B9市出租车（的士...</td>\n",
       "      <td>B9市出租车（的士）随口说价，不打表和提供正规发票</td>\n",
       "      <td>客运 管理处 出租车 城市 文明 窗口 市民 交通 出行 保障 出租车 的士 予以 出租车 ...</td>\n",
       "      <td>[客运, 管理处, 出租车, 城市, 文明, 窗口, 市民, 交通, 出行, 保障, 出租车...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>160</td>\n",
       "      <td>L9县县汽车站门口的右转弯车道成出租车专用停车道了</td>\n",
       "      <td>2019/7/24 21:23:31</td>\n",
       "      <td>L9县汽车站门口交通要道一条右转弯一条直行车道，如今右转弯车道却成了出租车的专用停车车道了，...</td>\n",
       "      <td>L9县县汽车站门口的右转弯车道成出租车专用停车道了</td>\n",
       "      <td>汽车站 门口 交通要道 转弯 行车道 转弯 车道 出租车 专用 停车 车道 直行 红绿灯 时...</td>\n",
       "      <td>[汽车站, 门口, 交通要道, 转弯, 行车道, 转弯, 车道, 出租车, 专用, 停车, ...</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                         title                 time  \\\n",
       "134                 请降低出租车行业门槛    2014/7/18 8:31:37   \n",
       "136                对A市出租车管理的建议  2011/11/18 15:48:58   \n",
       "138            我建议A市出租车最好搞个体经营   2013/3/23 20:52:51   \n",
       "140            请书记为M4市出租车司机做主！    2014/3/2 20:09:32   \n",
       "141           请问G市网约车政策何时出台和落实   2017/7/12 11:39:24   \n",
       "142      网约车无序，A市出租车司机的权益谁来保障？   2019/7/15 20:11:28   \n",
       "144  B9市出租车（的士）随口说价，不打表和提供正规发票    2015/6/28 9:26:35   \n",
       "160  L9县县汽车站门口的右转弯车道成出租车专用停车道了   2019/7/24 21:23:31   \n",
       "\n",
       "                                               content  \\\n",
       "134  出租车司机收入水平下降，黑车越来越多。不利于人民富足，而且增加政府行政成本。解决问题的关键在...   \n",
       "136  作为一个省级政府机关的公务员，近二十年来走遍中国的大城市，只有A市的出租车没有一点规矩，最不...   \n",
       "138  李总理讲：凡是市场化能解决的，都要由市场决定，政府不要插手（大意）。又讲：触动利益比触动灵魂...   \n",
       "140  我是一名普通的出租车司机，出生在一个普通的工人家中，一家五口就靠父亲那点微薄的工资来维持生活...   \n",
       "141  我是一名出租车司机，从事出租车行业已经有八年的时间了，由于网约车的兴起，现在出租车的收入大不...   \n",
       "142  黑车猖狂，网约车无序，的士夹缝中求生存，A市一有大事就狠抓合规出租车，可根本问题却是出租车司...   \n",
       "144  B9市客运管理处：出租车是城市文明的窗口，市民交通出行最基本的保障，现针对B9市出租车（的士...   \n",
       "160  L9县汽车站门口交通要道一条右转弯一条直行车道，如今右转弯车道却成了出租车的专用停车车道了，...   \n",
       "\n",
       "                        title_  \\\n",
       "134                 请降低出租车行业门槛   \n",
       "136                对A市出租车管理的建议   \n",
       "138            我建议A市出租车最好搞个体经营   \n",
       "140            请书记为M4市出租车司机做主！   \n",
       "141           请问G市网约车政策何时出台和落实   \n",
       "142      网约车无序，A市出租车司机的权益谁来保障？   \n",
       "144  B9市出租车（的士）随口说价，不打表和提供正规发票   \n",
       "160  L9县县汽车站门口的右转弯车道成出租车专用停车道了   \n",
       "\n",
       "                                              content_  \\\n",
       "134  出租车 司机 收入水平 下降 黑车 不利于 人民富足 增加 政府 行政 成本 解决问题 关键...   \n",
       "136  政府 机关 公务员 走遍 中国 大城市 出租车 规矩 文明 出游 乘兴 出租车 机场 火车站...   \n",
       "138  总理 市场化 解决 市场 政府 插手 触动 利益 触动 灵魂 出租车 个体经营 驾驶员 交警...   \n",
       "140  出租车 司机 出生 工人 家中 父亲 工资 生活 家中 无田无土 借钱 出租车 不到 时间 ...   \n",
       "141  出租车 司机 出租车 行业 八年 时间 网约车 兴起 出租车 收入 大不如前 收入 保安 工...   \n",
       "142          黑车 网约 无序 的士 夹缝 求生存 狠抓 合规 出租车 出租车 司机 权益 保障   \n",
       "144  客运 管理处 出租车 城市 文明 窗口 市民 交通 出行 保障 出租车 的士 予以 出租车 ...   \n",
       "160  汽车站 门口 交通要道 转弯 行车道 转弯 车道 出租车 专用 停车 车道 直行 红绿灯 时...   \n",
       "\n",
       "                                           content_cut  label  rank  \\\n",
       "134  [出租车, 司机, 收入水平, 下降, 黑车, 不利于, 人民富足, 增加, 政府, 行政,...      0     1   \n",
       "136  [政府, 机关, 公务员, 走遍, 中国, 大城市, 出租车, 规矩, 文明, 出游, 乘兴...      0     1   \n",
       "138  [总理, 市场化, 解决, 市场, 政府, 插手, 触动, 利益, 触动, 灵魂, 出租车,...      0     1   \n",
       "140  [出租车, 司机, 出生, 工人, 家中, 父亲, 工资, 生活, 家中, 无田无土, 借钱...      0     1   \n",
       "141  [出租车, 司机, 出租车, 行业, 八年, 时间, 网约车, 兴起, 出租车, 收入, 大...      0     1   \n",
       "142  [黑车, 网约, 无序, 的士, 夹缝, 求生存, 狠抓, 合规, 出租车, 出租车, 司机...      0     1   \n",
       "144  [客运, 管理处, 出租车, 城市, 文明, 窗口, 市民, 交通, 出行, 保障, 出租车...      0     1   \n",
       "160  [汽车站, 门口, 交通要道, 转弯, 行车道, 转弯, 车道, 出租车, 专用, 停车, ...      0     1   \n",
       "\n",
       "                                                matrix  \n",
       "134  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...  \n",
       "136  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...  \n",
       "138  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...  \n",
       "140  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...  \n",
       "141  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...  \n",
       "142  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...  \n",
       "144  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...  \n",
       "160  [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, ...  "
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 查看最大的类\n",
    "df_non_outliers[df_non_outliers['rank'] == 1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "data_pca_tsne.shape= (20, 2)\n"
     ]
    }
   ],
   "source": [
    "data_pca_tsne = feature_reduction(df_non_outliers['matrix'].tolist(),pca_n_components=3, tsne_n_components=2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "D:\\Anaconda3\\lib\\site-packages\\matplotlib\\backends\\backend_agg.py:211: RuntimeWarning: Glyph 8722 missing from current font.\n",
      "  font.set_text(s, 0.0, flags=flags)\n",
      "D:\\Anaconda3\\lib\\site-packages\\matplotlib\\backends\\backend_agg.py:180: RuntimeWarning: Glyph 8722 missing from current font.\n",
      "  font.set_text(s, 0, flags=flags)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 画个密度图\n",
    "import matplotlib.pyplot as plt\n",
    "plt.rcParams['font.family'] = ['sans-serif']\n",
    "plt.rcParams['font.sans-serif'] = ['SimHei']\n",
    "\n",
    "df_non_outliers['pca_tsne'] = data_pca_tsne.tolist()\n",
    "del df_non_outliers['matrix']\n",
    "data_pca_tsne = df_non_outliers['pca_tsne']\n",
    "label = df_non_outliers['label']\n",
    "\n",
    "plt.figure()\n",
    "x = [i[0] for i in data_pca_tsne]\n",
    "y = [i[1] for i in data_pca_tsne]\n",
    "plt.scatter(x, y, c=label)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1300x600 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1300x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def draw_clustering_analysis_barh(rank_num, value, yticks, title):\n",
    "    \"\"\"绘制聚类分析结果条形图\"\"\"\n",
    "    plt.figure(figsize=(13, 6), dpi=100)\n",
    "    plt.subplot(122)\n",
    "    ax = plt.gca()\n",
    "    ax.spines['left'].set_visible(False)\n",
    "    ax.spines['right'].set_visible(False)\n",
    "    ax.spines['top'].set_visible(False)\n",
    "    ax.invert_yaxis()\n",
    "    plt.barh(range(1, rank_num + 1), value, align='center', linewidth=0)\n",
    "    plt.yticks(range(1, rank_num + 1), yticks)\n",
    "    for a, b in zip(value, range(1, rank_num + 1)):\n",
    "        plt.text(a + 1, b, '%.0f' % a, ha='left', va='center')\n",
    "    plt.title(title)\n",
    "    plt.savefig( '2.jpg')\n",
    "    plt.show()\n",
    "    \n",
    "def draw_clustering_analysis_pie(rank_num, value, yticks,title):\n",
    "    \"\"\"绘制聚类分析结果饼图\"\"\"\n",
    "    plt.figure(figsize=(13, 6), dpi=100)\n",
    "    plt.subplot(132)\n",
    "    plt.pie(value, explode=[0.2] * rank_num, labels=yticks, autopct='%1.2f%%', pctdistance=0.7)\n",
    "    plt.title(title)\n",
    "    plt.show()\n",
    "\n",
    "    \n",
    "\n",
    "rank_num = get_num_of_value_no_repeat(df_non_outliers['rank'])\n",
    "value = [df_non_outliers[df_non_outliers['rank'] == i].shape[0] for i in range(1, rank_num + 1)]\n",
    "yticks1 = [str(get_most_common_words(df_non_outliers[df_non_outliers['rank'] == i]['content_cut'],\n",
    "                                             top_n=10)) + str(i) for i in range(1, rank_num + 1)]\n",
    "\n",
    "\n",
    "draw_clustering_analysis_barh(rank_num, value, yticks1, title='热点新闻分布条形图')\n",
    "draw_clustering_analysis_pie(rank_num, value, yticks1,title='热点新闻分布饼图')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}