{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "A9503F74C90B44568BD45B72D2C0E1A3",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "## 一、项目背景\n",
    "CDNow曾经是一家在线音乐零售平台，后被德国波泰尔斯曼娱乐集团公司出资收购，其资产总价值在最辉煌时曾超过10亿美元。本文主要通过分析CDNow网站的用户购买明细来分析该网站的用户消费行为，使运营部门在营销时更加具有针对性，从而节省成本，提升效率。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "D9DD7FB1882D40B18D004A3138F75C2E",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "## 二、提出问题\n",
    "对用户进行消费特征分析。分析框架如下：\n",
    "\n",
    "![Image Name](https://cdn.kesci.com/upload/image/q6z13o2jy1.png?imageView2/0/w/960/h/960)\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "E19AB080EDA9455AB4FD01912A3286B7",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "## 三.数据处理\n",
    "### 3.1 导入数据"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "id": "D16AD0E36E1D428ABF9C47170B0B8790",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "#导入常用的库\n",
    "import pandas as pd \n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "from datetime import datetime\n",
    "plt.style.use('ggplot')  #更改设计风格，使用自带的形式进行美化，这是一个r语言的风格"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "id": "F405E776A0084332A43D023B3605A426",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "#导入源数据\n",
    "columns = ['用户ID','购买日期','订单数','订单金额']\n",
    "df = pd.read_table(\"d:/CDNOW.txt\",names = columns,sep = '\\s+')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "11C982569CE347919E11A01D055405B6",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 加载包和数据，文件是txt，用read_table方法打开，因为原始数据不包含表头，所以需要赋予。字符串是空格分割，用\\s+表示匹配任意空白符。\n",
    "* 一般csv的数据分隔是以逗号的形式，但是这份数据它是通过多个空格来进行分隔\n",
    "* 消费行业或者是电商行业一般是通过订单数，订单额，购买日期，用户ID这四个字段来分析的。基本上这四个字段就可以进行很丰富的分析。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "id": "5C2054B80BFB4B0B8C09B724C7BE61C9",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "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>用户ID</th>\n",
       "      <th>购买日期</th>\n",
       "      <th>订单数</th>\n",
       "      <th>订单金额</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>19970101</td>\n",
       "      <td>1</td>\n",
       "      <td>11.77</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>19970112</td>\n",
       "      <td>1</td>\n",
       "      <td>12.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>19970112</td>\n",
       "      <td>5</td>\n",
       "      <td>77.00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>19970102</td>\n",
       "      <td>2</td>\n",
       "      <td>20.76</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>19970330</td>\n",
       "      <td>2</td>\n",
       "      <td>20.76</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   用户ID      购买日期  订单数   订单金额\n",
       "0     1  19970101    1  11.77\n",
       "1     2  19970112    1  12.00\n",
       "2     2  19970112    5  77.00\n",
       "3     3  19970102    2  20.76\n",
       "4     3  19970330    2  20.76"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#默认输出前五行\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "CAD805A5FACA40A78633123C87C7E8F9",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 观察数据，购买日期列表示时间，但现在它只是年月日组合的一串数字，数据不是时间类型，没有时间含义，需要转换。购买金额是小数。\n",
    "* 数据中存在一个用户在同一天或不同天下多次订单的情况，如用户ID为2的用户就在1月12日买了两次。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "B6CE10A71E054D3A88CCD9EC460234D1",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "### 3.2 描述性统计"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "id": "986623EF88EC49758DC23D1C1E773688",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "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>用户ID</th>\n",
       "      <th>购买日期</th>\n",
       "      <th>订单数</th>\n",
       "      <th>订单金额</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>69659.000000</td>\n",
       "      <td>6.965900e+04</td>\n",
       "      <td>69659.000000</td>\n",
       "      <td>69659.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>11470.854592</td>\n",
       "      <td>1.997228e+07</td>\n",
       "      <td>2.410040</td>\n",
       "      <td>35.893648</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>6819.904848</td>\n",
       "      <td>3.837735e+03</td>\n",
       "      <td>2.333924</td>\n",
       "      <td>36.281942</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>1.997010e+07</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>5506.000000</td>\n",
       "      <td>1.997022e+07</td>\n",
       "      <td>1.000000</td>\n",
       "      <td>14.490000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>11410.000000</td>\n",
       "      <td>1.997042e+07</td>\n",
       "      <td>2.000000</td>\n",
       "      <td>25.980000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>17273.000000</td>\n",
       "      <td>1.997111e+07</td>\n",
       "      <td>3.000000</td>\n",
       "      <td>43.700000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>23570.000000</td>\n",
       "      <td>1.998063e+07</td>\n",
       "      <td>99.000000</td>\n",
       "      <td>1286.010000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "               用户ID          购买日期           订单数          订单金额\n",
       "count  69659.000000  6.965900e+04  69659.000000  69659.000000\n",
       "mean   11470.854592  1.997228e+07      2.410040     35.893648\n",
       "std     6819.904848  3.837735e+03      2.333924     36.281942\n",
       "min        1.000000  1.997010e+07      1.000000      0.000000\n",
       "25%     5506.000000  1.997022e+07      1.000000     14.490000\n",
       "50%    11410.000000  1.997042e+07      2.000000     25.980000\n",
       "75%    17273.000000  1.997111e+07      3.000000     43.700000\n",
       "max    23570.000000  1.998063e+07     99.000000   1286.010000"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#数值列的汇总统计信息\n",
    "df.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0A594E6F029349F3A6150B15453AA326",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "describe是描述统计，对用户数据特征进行整体性判断：\n",
    "\n",
    "* 从数据的统计描述信息中可以看出，用户每个订单平均购买2.41个商品，每个订单平均消费35.89元。\n",
    "* 购买商品数量的标准差为2.33，说明数据具有一定的波动性；中位数为2个商品,75分位数为3个商品，说明大部分订单的购买数量都不多。最大值在99个，数字比较高。购买金额的情况差不多，大部分订单都集中在小额。\n",
    "* 一般而言，消费类的数据分布，都是长尾形态。大部分用户都是小额，然而小部分用户贡献了收入的大头，俗称二八。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "60BD57422E8C42229E7C92008E1CEB17",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "## 3.3 数据处理"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "id": "66B5E06F232344498F9DCFF5EE1B2E69",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "RangeIndex: 69659 entries, 0 to 69658\n",
      "Data columns (total 4 columns):\n",
      "用户ID    69659 non-null int64\n",
      "购买日期    69659 non-null int64\n",
      "订单数     69659 non-null int64\n",
      "订单金额    69659 non-null float64\n",
      "dtypes: float64(1), int64(3)\n",
      "memory usage: 2.1 MB\n"
     ]
    }
   ],
   "source": [
    "#索引，数据类型和内存信息\n",
    "df.info()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "CD763CA5D33C4C6FA807758468DF106B",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "查看数据类型、数据是否存在空值；原数据没有空值，很干净的数据。接下来我们要将时间的数据类型转化。\n",
    "当利用pandas进行数据处理的时候，经常会遇见数据类型的问题，当拿到数据的时候，首先要确定拿到的是正确的数据类型，如果数据类型不正确需要进行数据类型的转化，再进行数据处理。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "id": "F92870493E94427180FE19099C907E26",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "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>用户ID</th>\n",
       "      <th>购买日期</th>\n",
       "      <th>订单数</th>\n",
       "      <th>订单金额</th>\n",
       "      <th>月份</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>1997-01-01</td>\n",
       "      <td>1</td>\n",
       "      <td>11.77</td>\n",
       "      <td>1997-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>1997-01-12</td>\n",
       "      <td>1</td>\n",
       "      <td>12.00</td>\n",
       "      <td>1997-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2</td>\n",
       "      <td>1997-01-12</td>\n",
       "      <td>5</td>\n",
       "      <td>77.00</td>\n",
       "      <td>1997-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>3</td>\n",
       "      <td>1997-01-02</td>\n",
       "      <td>2</td>\n",
       "      <td>20.76</td>\n",
       "      <td>1997-01-01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3</td>\n",
       "      <td>1997-03-30</td>\n",
       "      <td>2</td>\n",
       "      <td>20.76</td>\n",
       "      <td>1997-03-01</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   用户ID       购买日期  订单数   订单金额         月份\n",
       "0     1 1997-01-01    1  11.77 1997-01-01\n",
       "1     2 1997-01-12    1  12.00 1997-01-01\n",
       "2     2 1997-01-12    5  77.00 1997-01-01\n",
       "3     3 1997-01-02    2  20.76 1997-01-01\n",
       "4     3 1997-03-30    2  20.76 1997-03-01"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 将购买日期列进行数据类型转换\n",
    "df['购买日期'] = pd.to_datetime(df.购买日期,format = '%Y%m%d') #Y四位数的日期部分，y表示两位数的日期部分\n",
    "df['月份'] = df.购买日期.values.astype('datetime64[M]') \n",
    "df.head()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "90C0526E76244E4A86EB3070AC4E69CC",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* pd.to_datetime可以将特定的字符串或者数字转换成时间格式，其中的format参数用于匹配。例如19970101，%Y匹配前四位数字1997，如果y小写只匹配两位数字97，%m匹配01，%d匹配01。\n",
    "* 另外，小时是%h，分钟是%M，注意和月的大小写不一致，秒是%s。若是1997-01-01这形式，则是%Y-%m-%d，以此类推。\n",
    "* astype也可以将时间格式进行转换，比如[M]转化成月份。我们将月份作为消费行为的主要事件窗口，选择哪种时间窗口取决于消费频率。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "00D373E65E10471DA4FF690C941AAFA3",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "## 四.用户总体消费趋势分析"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "id": "ED880AA60B374AF79F7E0DDD28160D45",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x1080 with 4 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 设置图的大小，添加子图\n",
    "from matplotlib.font_manager import FontProperties\n",
    "font1 = FontProperties(fname=r\"c:/windows/fonts/simhei.ttf\")\n",
    "plt.figure(figsize=(20,15))\n",
    "# 每月的总销售额\n",
    "plt.subplot(221)\n",
    "df.groupby('月份')['订单金额'].sum().plot(fontsize=24) \n",
    "plt.title('总销售额',fontproperties=font1,fontsize=24) \n",
    " \n",
    "#每月的消费次数\n",
    "plt.subplot(222)\n",
    "df.groupby('月份')['购买日期'].count().plot(fontsize=24)\n",
    "plt.title('消费次数',fontproperties=font1,fontsize=24) \n",
    "\n",
    "#每月的销量\n",
    "plt.subplot(223)\n",
    "df.groupby('月份')['订单数'].sum().plot(fontsize=24)\n",
    "plt.title('总销量',fontsize=24)\n",
    "\n",
    "#每月的消费人数\n",
    "plt.subplot(224)\n",
    "df.groupby('月份')['用户ID'].apply(lambda x:len(x.unique())).plot(fontsize=24)\n",
    "plt.title('消费人数',fontsize=24)\n",
    "plt.tight_layout() # 设置子图的间距\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "25A412CCB14244D08C315F73BCB5B8A7",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 四个折线图的整体趋势基本一致，可以看出来，1997年前3个月的销量特别高，随之而来的销售额也是暴涨，在3月份之后骤然下降，接近平稳。\n",
    "* 为什么会呈现这个原因呢？我们假设是用户身上出了问题，早期时间段的用户中有异常值，第二假设是各类促销营销，但这里只有消费数据，所以无法判断。\n",
    "* 另一方面，在2月到3月这段期间，可以发现消费人数稍有下降，但总销量与总销售额却依然上升，是不是说明3月份的用户中有我们需要重点发展的高价值客户呢？"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "2A0B10882C49417887B47142EEF8AE1B",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "## 五. 用户个体消费数据分析\n",
    "### 5.1 用户消费金额，消费次数的描述统计"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "id": "395D98A6450B400A81EBDE98E4273E13",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "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>订单数</th>\n",
       "      <th>订单金额</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>count</th>\n",
       "      <td>23570.000000</td>\n",
       "      <td>23570.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>mean</th>\n",
       "      <td>7.122656</td>\n",
       "      <td>106.080426</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>std</th>\n",
       "      <td>16.983531</td>\n",
       "      <td>240.925195</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>min</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25%</th>\n",
       "      <td>1.000000</td>\n",
       "      <td>19.970000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50%</th>\n",
       "      <td>3.000000</td>\n",
       "      <td>43.395000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75%</th>\n",
       "      <td>7.000000</td>\n",
       "      <td>106.475000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>max</th>\n",
       "      <td>1033.000000</td>\n",
       "      <td>13990.930000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                订单数          订单金额\n",
       "count  23570.000000  23570.000000\n",
       "mean       7.122656    106.080426\n",
       "std       16.983531    240.925195\n",
       "min        1.000000      0.000000\n",
       "25%        1.000000     19.970000\n",
       "50%        3.000000     43.395000\n",
       "75%        7.000000    106.475000\n",
       "max     1033.000000  13990.930000"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 根据用户id进行分组\n",
    "group_user = df.groupby('用户ID').sum()\n",
    "group_user.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0A47EF4FC9C94704ABE8F1C2CB40AFA0",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "从用户角度看，每位用户平均购买7张CD，最多的用户购买了1033张。用户的平均消费金额（客单价）100元，标准差是240，结合分位数和最大值看，平均值才和75分位接近，肯定存在小部分的高额消费用户,这也符合二八法则。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "CE7C9556DBE94B5FB03B5DEB1C689823",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "### 5.2 用户消费金额和消费次数的散点图"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "id": "F56F92333FE942CE82F98082E80C4CAC",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0xeecc230>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "#查询条件：订单金额 < 4000\n",
    "group_user.query('订单金额 < 4000').plot.scatter(x='订单金额',y='订单数')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "CD6152377EBA4E2AA2D6E4D44AA7A129",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 绘制用户的散点图，用户比较健康而且规律性很强。因为这是CD网站的销售数据，商品比较单一，金额和商品量的关系也因此呈线性，没几个离群点。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0CA21CACCAE944ED8C647BE60C080E98",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "### 5.3 用户消费金额的分布图（二八法则）"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "id": "C024EC1533AC49E28C62683CEE583356",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f94c5ecbba8>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"https://cdn.kesci.com/rt_upload/C024EC1533AC49E28C62683CEE583356/q6z2ki8vm7.png\">"
      ],
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "group_user.订单金额. plot.hist(bins = 20)\n",
    "#bins = 20,就是分成20块，最高金额是14000，每个项就是700"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "FAFF5B4D277A4B889263077B1D04AB26",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 从图上看出，用户的消费呈集中趋势，可能是有个别的极大值干扰导致。\n",
    "* 可以排除极大值再看看分布情况"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "id": "18E9CB3442A3492791D27E822308FAA2",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f94c5b2e390>"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"https://cdn.kesci.com/rt_upload/18E9CB3442A3492791D27E822308FAA2/q6z2msr7ov.png\">"
      ],
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "group_user.query(\"订单金额< 800\")['订单金额'].plot.hist(bins=20)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "32ABD1CC2BDC4818AC6A64133C7B47A8",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "筛选出了消费金额小于800的用户，我们可以看到：\n",
    "\n",
    "* 大部分用户的消费能力并不高，将近半数的用户消费金额不超过40元，高消费用户（ >200元 ）不超过2000人。\n",
    "* 从上图直方图可知，大部分用户的消费能力确实不高，绝大部分呈现集中在很低的消费档次。高消费用户在图上几乎看不到，这也确实符合消费行为的行业规律。\n",
    "* 虽然有极端数据干扰，但是大部分的用户还是集中在比较低的消费档次。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "1F57D8FD014344B882F7F24528EC6669",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "### 5.4 用户消费次数的分布图（二八法则）"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "id": "DB907FF2120847BFB8EF55C59081C1D7",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f94e8afe6d8>"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"https://cdn.kesci.com/rt_upload/DB907FF2120847BFB8EF55C59081C1D7/q6z2s44l6x.png\">"
      ],
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "group_user.query('订单数 < 100').订单数.hist(bins = 40)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "92160C87EE56402F9AE8698267C629AB",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 大部分用户购买CD的数量都是在3张以内，购买大量CD的用户数量并不多。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4DF7E5E80C004CA098470F27EA695739",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "## 六.用户消费周期分析\n",
    "### 6.1 用户购买周期（按订单）"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "id": "447B3F8EDF5F43668CBDAC3F9D31BF2A",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "用户ID   \n",
       "1     0        NaT\n",
       "2     1        NaT\n",
       "      2     0 days\n",
       "3     3        NaT\n",
       "      4    87 days\n",
       "      5     3 days\n",
       "      6   227 days\n",
       "      7    10 days\n",
       "      8   184 days\n",
       "4     9        NaT\n",
       "Name: 购买日期, dtype: timedelta64[ns]"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#每个用户的每次购买时间间隔\n",
    "order_diff = df.groupby('用户ID').apply(lambda x:x['购买日期'] - x['购买日期'].shift())\n",
    "order_diff.head(10)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "id": "A09471A4E12945F78FDA6050DE25BFD3",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "count                      46089\n",
       "mean     68 days 23:22:13.567662\n",
       "std      91 days 00:47:33.924168\n",
       "min              0 days 00:00:00\n",
       "25%             10 days 00:00:00\n",
       "50%             31 days 00:00:00\n",
       "75%             89 days 00:00:00\n",
       "max            533 days 00:00:00\n",
       "Name: 购买日期, dtype: object"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "order_diff.describe()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "7895B31997F04BD19A6CF402AE542060",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 平均每个用户的购买时间间隔是68天，间隔最长的是533天。想要召回用户，在60天左右的消费间隔是比较好的。\n",
    "* 绝大部分用户的消费周期都低于100天。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "17024504F0F64460878DD54491D61620",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "### 6.1.2 用户消费周期分布"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "id": "E5200581D402419C820094D52C983493",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://cdn.kesci.com/rt_upload/E5200581D402419C820094D52C983493/q6z2ymlsvc.png\">"
      ],
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(15,5))\n",
    "plt.hist((order_diff / np.timedelta64(1, 'D')).dropna(), bins = 50)\n",
    "plt.xlabel('消费周期',fontsize=24)\n",
    "plt.ylabel('频数',fontsize=24)\n",
    "plt.title('用户消费周期分布图',fontsize=24);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "5906292A167C4EA4B64BAE35ED6EB338",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 典型的（指数）长尾分布，大部分用户的消费间隔确实比较短。不妨将时间召回点设为消费后立即赠送优惠券，消费后10天询问用户礼品怎么样，消费后20天提醒优惠券到期，消费后30天短信推送。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "35EC72DE2111480D84178545E1FE392D",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "### 6.2 用户生命周期（按第一次&最后一次消费）"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "id": "851ED6729C164029B703974F2B789D8E",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "用户ID\n",
       "1     0 days\n",
       "2     0 days\n",
       "3   511 days\n",
       "4   345 days\n",
       "5   367 days\n",
       "Name: 购买日期, dtype: timedelta64[ns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "orderdt_min=df.groupby('用户ID').购买日期.min()#第一次消费\n",
    "orderdt_max=df.groupby('用户ID').购买日期.max()#最后一次消费\n",
    "(orderdt_max-orderdt_min).head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "id": "61BC4E3074594E3B8EBA41D549A5AA31",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timedelta('134 days 20:55:36.987696')"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#计算用户的平均生命周期\n",
    "(orderdt_max-orderdt_min).mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "35D84C50ACBF4B7D8E75DCDC9B99BEE2",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 所有用户的平均生命周期是134天，比预想的高，但是平均数不具有代表性，接下来还是看一下分布情况。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "id": "60037F90C030427B8742D71CF208A017",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f94c59044a8>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"https://cdn.kesci.com/rt_upload/60037F90C030427B8742D71CF208A017/q6z36ue1vk.png\">"
      ],
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "((orderdt_max-orderdt_min)/np.timedelta64(1,'D')).hist(bins=15)\n",
    "#因为数据类型是timedelta时间，无法直接作出直方图，所以先换算成数值。\n",
    "#换算的方式直接除timedelta函数即可，np.timedelta64(1, ‘D’)，D表示天，1表示1天，作为单位使用的。\n",
    "#因为max-min已经表示为天了，两者相除就是周期"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "1162CE82B9AE46D1833940FD95609AEE",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 大部分用户只消费了一次，所有生命周期的大头都集中在了0天。\n",
    "* 但这不是我们想要的答案，不妨将只消费了一次的新客排除，来计算所有消费过两次以上的老客的生命周期。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "id": "C53EA6575C0F4D6ABA0D31CB084D5034",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "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>用户ID</th>\n",
       "      <th>购买日期</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>0 days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2</td>\n",
       "      <td>0 days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>3</td>\n",
       "      <td>511 days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4</td>\n",
       "      <td>345 days</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>5</td>\n",
       "      <td>367 days</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "   用户ID     购买日期\n",
       "0     1   0 days\n",
       "1     2   0 days\n",
       "2     3 511 days\n",
       "3     4 345 days\n",
       "4     5 367 days"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#计算所有消费过两次以上的老客的生命周期\n",
    "life_time = (orderdt_max - orderdt_min).reset_index()\n",
    "life_time.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "id": "AD7770867D3843918FD001A1C9D8A452",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7f94c5852e48>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"https://cdn.kesci.com/rt_upload/AD7770867D3843918FD001A1C9D8A452/q6z39p2y8z.png\">"
      ],
      "text/plain": [
       "<Figure size 864x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#用户生命周期分布图\n",
    "plt.figure(figsize=(10,5))\n",
    "life_time['life_time'] = life_time.购买日期 / np.timedelta64(1,'D')\n",
    "life_time[life_time.life_time > 0].life_time.hist(bins = 100, figsize = (12,6))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "B2824B6BE06C40FD9F190D56DF22064C",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 上图可见，用户生命周期呈现双峰趋势，20天内生命周期的用户是一个高峰，400至500天内生命周期的用户是另一个高峰。\n",
    "* 根据此情况，应该在20天内对客户进行引导，促进其再次消费并形成消费习惯，延长其生命周期；在100至400天的用户，也要根据其特点推出有针对性的营销活动，引导其持续消费。\n",
    "分析去掉0天生命周期的用户之后的用户生命周期的平均值。\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "id": "EA4DFAEE5E0F456E9C09AD5B453B0F32",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Timedelta('276 days 01:04:31.344216')"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#去掉0天生命周期的用户之后的用户生命周期的平均值\n",
    "life_time[life_time.life_time>0].购买日期.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "303A170A83894DD6BB5D1E7C2766E61B",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 可见，若在用户首次消费后，加强对其再次消费的引导，可将其生命周期延长至原来的两倍。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "6E986B7EBE9C4DFD83293A15D24D6548",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "## 七.用户分层\n",
    "按用户价值分层---RFM模型\n",
    "为了进行精细化运营，可以利用RMF模型对用户价值指数（衡量历史到当前用户贡献的收益）进行计算，其中\n",
    "最近一次消费-R：客户最近一次交易时间的间隔。R值越大，表示客户交易发生的日期越久，反之则交易发生的日期越近。\n",
    "消费频率-F：客户在最近一段时间内交易的次数。F值越大，表示客户交易越频繁，反之则表示客户交易不够活跃。\n",
    "消费金额-M：客户在最近一段时间内交易的金额。M值越大，表示客户价值越高，反之则表示客户价值越低。\n",
    "根据上述三个维度，对客户做细分"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "id": "EE077FB03BF244529869E47EB01BAEAF",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "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>F</th>\n",
       "      <th>M</th>\n",
       "      <th>R</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>label</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>一般价值客户</th>\n",
       "      <td>650</td>\n",
       "      <td>7181.28</td>\n",
       "      <td>36295.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>一般保持客户</th>\n",
       "      <td>1712</td>\n",
       "      <td>19937.45</td>\n",
       "      <td>29448.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>一般发展客户</th>\n",
       "      <td>13977</td>\n",
       "      <td>196971.23</td>\n",
       "      <td>591108.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>一般挽留客户</th>\n",
       "      <td>29346</td>\n",
       "      <td>438291.81</td>\n",
       "      <td>6951815.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>重要价值客户</th>\n",
       "      <td>11121</td>\n",
       "      <td>167080.83</td>\n",
       "      <td>358363.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>重要保持客户</th>\n",
       "      <td>107789</td>\n",
       "      <td>1592039.62</td>\n",
       "      <td>517267.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>重要发展客户</th>\n",
       "      <td>2023</td>\n",
       "      <td>45785.01</td>\n",
       "      <td>56636.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>重要挽留客户</th>\n",
       "      <td>1263</td>\n",
       "      <td>33028.40</td>\n",
       "      <td>114482.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "             F           M          R\n",
       "label                                \n",
       "一般价值客户     650     7181.28    36295.0\n",
       "一般保持客户    1712    19937.45    29448.0\n",
       "一般发展客户   13977   196971.23   591108.0\n",
       "一般挽留客户   29346   438291.81  6951815.0\n",
       "重要价值客户   11121   167080.83   358363.0\n",
       "重要保持客户  107789  1592039.62   517267.0\n",
       "重要发展客户    2023    45785.01    56636.0\n",
       "重要挽留客户    1263    33028.40   114482.0"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfm = df.pivot_table(index = '用户ID',\n",
    "                     values = ['订单金额','购买日期','订单数'],\n",
    "                     aggfunc = {'订单金额':'sum',\n",
    "                                '购买日期':'max',\n",
    "                                '订单数':'sum'})\n",
    "rfm.head()\n",
    "\n",
    "# 日期的最大值与当前日期的差值为R\n",
    "rfm['R'] = (rfm['购买日期'].max() - rfm['购买日期']) / np.timedelta64(1,'D')\n",
    "rfm.rename(columns = {'订单金额':'M',\n",
    "                     '订单数':'F'},\n",
    "          inplace=True)\n",
    "\n",
    "# 构建rfm模型公式\n",
    "def get_rfm(x):\n",
    "    level = x.apply(lambda x:'1' if x>=0 else '0')\n",
    "    label = level['R'] + level['F'] + level['M']\n",
    "    d = {'111':'重要价值客户',\n",
    "        '011':'重要保持客户',\n",
    "        '101':'重要挽留客户',\n",
    "        '001':'重要发展客户',\n",
    "        '110':'一般价值客户',\n",
    "        '010':'一般保持客户',\n",
    "        '100':'一般挽留客户',\n",
    "        '000':'一般发展客户'}\n",
    "    \n",
    "    result = d[label]\n",
    "    return result\n",
    "rfm['label'] = rfm[['R','F','M']].apply(lambda x:(x-x.mean()) / x.std()).apply(get_rfm,axis=1)\n",
    "rfm.head()\n",
    "#求和\n",
    "rfm.groupby('label').sum()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "161E2440047B4D19874EFFD8BA564262",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* M列中不同层次客户的消费累计金额，重要保持客户的累计消费金额为159203.62，排名最高"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "id": "A4F0CC9D9AD0404280F6C0A6F57BD2BA",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "label\n",
       "一般价值客户       77\n",
       "一般保持客户      206\n",
       "一般发展客户     3300\n",
       "一般挽留客户    14074\n",
       "重要价值客户      787\n",
       "重要保持客户     4554\n",
       "重要发展客户      331\n",
       "重要挽留客户      241\n",
       "dtype: int64"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "rfm.groupby('label').size()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "4742CD4333274DBD8C0E398A13BCC8F7",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 以上为不同层次用户的消费人数，一般挽留用户的消费人数排名第一，有14074人，重要保持客户排名第二，有4554人，与一般挽留用户差距比较大，但累计消费金额最多，业务方可以根据结果对客户分类运营，降低营销成本，提高ROI。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "9C88C38427BB41F1B8AEDD0877CB34CB",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "## 八.用户质量分析"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "id": "70DBC91E1CAD45379918E4538F311BA3",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7f94c59365c0>"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"https://cdn.kesci.com/rt_upload/70DBC91E1CAD45379918E4538F311BA3/q6z3u7c99t.png\">"
      ],
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "a = df.groupby('用户ID')['购买日期'].agg(['min','max']).reset_index()\n",
    "new_old = (a['min'] == a['max']).value_counts().values\n",
    "\n",
    "plt.pie(x = new_old,\n",
    "       autopct = '%.1f%%',\n",
    "       shadow = True,\n",
    "       explode = [0.08,0],\n",
    "       textprops = {'fontsize' : 11})\n",
    "plt.axis('equal') \n",
    "plt.legend(['仅消费一次','多次消费'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "B38FB2330DF44E518C1C4DD64F04B14C",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 有超过一半的用户仅消费了一次，这也说明了运营不利，留存效果不好。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "0AF08A0145234843857D8D6C166D49CB",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 复购率的定义：在某时间窗口内消费两次及以上的用户在总消费用户中占比。这里的时间窗口是月，如果一个用户在同一天下了两笔订单，这里也将他算作复购用户"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "id": "7D4FFFA42442452F859A8DAC76070B7D",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "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>月份</th>\n",
       "      <th>1997-01-01 00:00:00</th>\n",
       "      <th>1997-02-01 00:00:00</th>\n",
       "      <th>1997-03-01 00:00:00</th>\n",
       "      <th>1997-04-01 00:00:00</th>\n",
       "      <th>1997-05-01 00:00:00</th>\n",
       "      <th>1997-06-01 00:00:00</th>\n",
       "      <th>1997-07-01 00:00:00</th>\n",
       "      <th>1997-08-01 00:00:00</th>\n",
       "      <th>1997-09-01 00:00:00</th>\n",
       "      <th>1997-10-01 00:00:00</th>\n",
       "      <th>1997-11-01 00:00:00</th>\n",
       "      <th>1997-12-01 00:00:00</th>\n",
       "      <th>1998-01-01 00:00:00</th>\n",
       "      <th>1998-02-01 00:00:00</th>\n",
       "      <th>1998-03-01 00:00:00</th>\n",
       "      <th>1998-04-01 00:00:00</th>\n",
       "      <th>1998-05-01 00:00:00</th>\n",
       "      <th>1998-06-01 00:00:00</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>用户ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>2.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "月份    1997-01-01  1997-02-01  1997-03-01  1997-04-01  1997-05-01  1997-06-01  \\\n",
       "用户ID                                                                           \n",
       "1            1.0         0.0         0.0         0.0         0.0         0.0   \n",
       "2            2.0         0.0         0.0         0.0         0.0         0.0   \n",
       "3            1.0         0.0         1.0         1.0         0.0         0.0   \n",
       "4            2.0         0.0         0.0         0.0         0.0         0.0   \n",
       "5            2.0         1.0         0.0         1.0         1.0         1.0   \n",
       "\n",
       "月份    1997-07-01  1997-08-01  1997-09-01  1997-10-01  1997-11-01  1997-12-01  \\\n",
       "用户ID                                                                           \n",
       "1            0.0         0.0         0.0         0.0         0.0         0.0   \n",
       "2            0.0         0.0         0.0         0.0         0.0         0.0   \n",
       "3            0.0         0.0         0.0         0.0         2.0         0.0   \n",
       "4            0.0         1.0         0.0         0.0         0.0         1.0   \n",
       "5            1.0         0.0         1.0         0.0         0.0         2.0   \n",
       "\n",
       "月份    1998-01-01  1998-02-01  1998-03-01  1998-04-01  1998-05-01  1998-06-01  \n",
       "用户ID                                                                          \n",
       "1            0.0         0.0         0.0         0.0         0.0         0.0  \n",
       "2            0.0         0.0         0.0         0.0         0.0         0.0  \n",
       "3            0.0         0.0         0.0         0.0         1.0         0.0  \n",
       "4            0.0         0.0         0.0         0.0         0.0         0.0  \n",
       "5            1.0         0.0         0.0         0.0         0.0         0.0  "
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#每个用户在每月的订单数\n",
    "pivoted_df=df.pivot_table(index='用户ID',columns='月份',values='购买日期',#pivot_table透视表\n",
    "                       aggfunc='count').fillna(0)#某些用户在某月没有消费过，用nan表示，这里用0填充\n",
    "\n",
    "pivoted_df.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "id": "C567A194B41B4ECA8FB7EAAAB03E0D72",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "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>月份</th>\n",
       "      <th>1997-01-01 00:00:00</th>\n",
       "      <th>1997-02-01 00:00:00</th>\n",
       "      <th>1997-03-01 00:00:00</th>\n",
       "      <th>1997-04-01 00:00:00</th>\n",
       "      <th>1997-05-01 00:00:00</th>\n",
       "      <th>1997-06-01 00:00:00</th>\n",
       "      <th>1997-07-01 00:00:00</th>\n",
       "      <th>1997-08-01 00:00:00</th>\n",
       "      <th>1997-09-01 00:00:00</th>\n",
       "      <th>1997-10-01 00:00:00</th>\n",
       "      <th>1997-11-01 00:00:00</th>\n",
       "      <th>1997-12-01 00:00:00</th>\n",
       "      <th>1998-01-01 00:00:00</th>\n",
       "      <th>1998-02-01 00:00:00</th>\n",
       "      <th>1998-03-01 00:00:00</th>\n",
       "      <th>1998-04-01 00:00:00</th>\n",
       "      <th>1998-05-01 00:00:00</th>\n",
       "      <th>1998-06-01 00:00:00</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>用户ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "月份    1997-01-01  1997-02-01  1997-03-01  1997-04-01  1997-05-01  1997-06-01  \\\n",
       "用户ID                                                                           \n",
       "1            0.0         NaN         NaN         NaN         NaN         NaN   \n",
       "2            1.0         NaN         NaN         NaN         NaN         NaN   \n",
       "3            0.0         NaN         0.0         0.0         NaN         NaN   \n",
       "4            1.0         NaN         NaN         NaN         NaN         NaN   \n",
       "5            1.0         0.0         NaN         0.0         0.0         0.0   \n",
       "\n",
       "月份    1997-07-01  1997-08-01  1997-09-01  1997-10-01  1997-11-01  1997-12-01  \\\n",
       "用户ID                                                                           \n",
       "1            NaN         NaN         NaN         NaN         NaN         NaN   \n",
       "2            NaN         NaN         NaN         NaN         NaN         NaN   \n",
       "3            NaN         NaN         NaN         NaN         1.0         NaN   \n",
       "4            NaN         0.0         NaN         NaN         NaN         0.0   \n",
       "5            0.0         NaN         0.0         NaN         NaN         1.0   \n",
       "\n",
       "月份    1998-01-01  1998-02-01  1998-03-01  1998-04-01  1998-05-01  1998-06-01  \n",
       "用户ID                                                                          \n",
       "1            NaN         NaN         NaN         NaN         NaN         NaN  \n",
       "2            NaN         NaN         NaN         NaN         NaN         NaN  \n",
       "3            NaN         NaN         NaN         NaN         0.0         NaN  \n",
       "4            NaN         NaN         NaN         NaN         NaN         NaN  \n",
       "5            0.0         NaN         NaN         NaN         NaN         NaN  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#转换：消费2次以上记为1，消费1次记为0，消费0次记为NAN  \n",
    "#applymap针对dataframe所有数据\n",
    "pivoted_df_transf=pivoted_df.applymap(lambda x: 1 if x>1 else np.nan if x==0 else 0)\n",
    "pivoted_df_transf.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "id": "3744E2C4CCA8406889D8968903994226",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5, 1.0, '复购率的变化')"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "text/html": [
       "<img src=\"https://cdn.kesci.com/rt_upload/3744E2C4CCA8406889D8968903994226/q6z3yvnx50.png\">"
      ],
      "text/plain": [
       "<Figure size 1080x432 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#count统计所有非空数据个数表示总消费用户数，sum计算非0数据的和表示消费两次以上的用户数\n",
    "df_duplicate =pd.DataFrame(pivoted_df_transf.sum()/pivoted_df_transf.count()).reset_index()\n",
    "df_duplicate.columns = ['Date', 'DuplicatedRate']\n",
    "df_duplicate['Date'] = df_duplicate.Date.astype(str).apply(lambda x:x[:-3])\n",
    "\n",
    "plt.figure(figsize = (15,6))\n",
    "plt.plot(df_duplicate.Date, df_duplicate.DuplicatedRate)\n",
    "plt.xlabel('时间', fontsize=24)\n",
    "plt.ylabel('复购率',fontsize=24)\n",
    "# plt.ylim(0,1)\n",
    "plt.title('复购率的变化',fontsize=24)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "BE53D4AC5AA4467A9BA138187C97F146",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 说明：图上可以看出复购率在早期，因为大量新用户加入的关系，新客的复购率并不高，如1月新客们的复购率只有6%左右。而在后期，这时的用户都是大浪淘沙剩下的老客，复购率比较稳定，在20%左右。单看新客和老客，复购率有三倍左右的差距。"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "10E831466FF14B3189D09590864467D0",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "回购率：是某一个时间窗口内消费的用户，在下一个时间窗口仍旧消费的占比。比如，我1月消费用户1000，他们中有300个2月依然消费，回购率是30%。"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "id": "AE9BB1544EDF4FF58E2719D724F91BFC",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "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>1997-01-01</th>\n",
       "      <th>1997-02-01</th>\n",
       "      <th>1997-03-01</th>\n",
       "      <th>1997-04-01</th>\n",
       "      <th>1997-05-01</th>\n",
       "      <th>1997-06-01</th>\n",
       "      <th>1997-07-01</th>\n",
       "      <th>1997-08-01</th>\n",
       "      <th>1997-09-01</th>\n",
       "      <th>1997-10-01</th>\n",
       "      <th>1997-11-01</th>\n",
       "      <th>1997-12-01</th>\n",
       "      <th>1998-01-01</th>\n",
       "      <th>1998-02-01</th>\n",
       "      <th>1998-03-01</th>\n",
       "      <th>1998-04-01</th>\n",
       "      <th>1998-05-01</th>\n",
       "      <th>1998-06-01</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>用户ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>11.77</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>44.50</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>20.76</td>\n",
       "      <td>0.0</td>\n",
       "      <td>20.76</td>\n",
       "      <td>19.54</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>39.205</td>\n",
       "      <td>0.000</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>16.99</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>29.53</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.00</td>\n",
       "      <td>14.96</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>26.480</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>21.65</td>\n",
       "      <td>38.9</td>\n",
       "      <td>0.00</td>\n",
       "      <td>45.55</td>\n",
       "      <td>38.71</td>\n",
       "      <td>26.14</td>\n",
       "      <td>28.14</td>\n",
       "      <td>0.00</td>\n",
       "      <td>40.47</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000</td>\n",
       "      <td>43.465</td>\n",
       "      <td>37.47</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.00</td>\n",
       "      <td>0.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      1997-01-01  1997-02-01  1997-03-01  1997-04-01  1997-05-01  1997-06-01  \\\n",
       "用户ID                                                                           \n",
       "1          11.77         0.0        0.00        0.00        0.00        0.00   \n",
       "2          44.50         0.0        0.00        0.00        0.00        0.00   \n",
       "3          20.76         0.0       20.76       19.54        0.00        0.00   \n",
       "4          29.53         0.0        0.00        0.00        0.00        0.00   \n",
       "5          21.65        38.9        0.00       45.55       38.71       26.14   \n",
       "\n",
       "      1997-07-01  1997-08-01  1997-09-01  1997-10-01  1997-11-01  1997-12-01  \\\n",
       "用户ID                                                                           \n",
       "1           0.00        0.00        0.00         0.0       0.000       0.000   \n",
       "2           0.00        0.00        0.00         0.0       0.000       0.000   \n",
       "3           0.00        0.00        0.00         0.0      39.205       0.000   \n",
       "4           0.00       14.96        0.00         0.0       0.000      26.480   \n",
       "5          28.14        0.00       40.47         0.0       0.000      43.465   \n",
       "\n",
       "      1998-01-01  1998-02-01  1998-03-01  1998-04-01  1998-05-01  1998-06-01  \n",
       "用户ID                                                                          \n",
       "1           0.00         0.0         0.0         0.0        0.00         0.0  \n",
       "2           0.00         0.0         0.0         0.0        0.00         0.0  \n",
       "3           0.00         0.0         0.0         0.0       16.99         0.0  \n",
       "4           0.00         0.0         0.0         0.0        0.00         0.0  \n",
       "5          37.47         0.0         0.0         0.0        0.00         0.0  "
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#回购率\n",
    "#每个用户每个月平均消费金额\n",
    "pivoted_money=df.pivot_table(index='用户ID',columns='月份',values='订单金额',\n",
    "                             aggfunc='mean').fillna(0)\n",
    "\n",
    "columns_month=df.月份.sort_values().astype('str').unique()\n",
    "pivoted_money.columns=columns_month\n",
    "pivoted_money.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "id": "34DB9E66FE2E4CEFBDC626D6516F320D",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "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>1997-01-01</th>\n",
       "      <th>1997-02-01</th>\n",
       "      <th>1997-03-01</th>\n",
       "      <th>1997-04-01</th>\n",
       "      <th>1997-05-01</th>\n",
       "      <th>1997-06-01</th>\n",
       "      <th>1997-07-01</th>\n",
       "      <th>1997-08-01</th>\n",
       "      <th>1997-09-01</th>\n",
       "      <th>1997-10-01</th>\n",
       "      <th>1997-11-01</th>\n",
       "      <th>1997-12-01</th>\n",
       "      <th>1998-01-01</th>\n",
       "      <th>1998-02-01</th>\n",
       "      <th>1998-03-01</th>\n",
       "      <th>1998-04-01</th>\n",
       "      <th>1998-05-01</th>\n",
       "      <th>1998-06-01</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>用户ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      1997-01-01  1997-02-01  1997-03-01  1997-04-01  1997-05-01  1997-06-01  \\\n",
       "用户ID                                                                           \n",
       "1              1           0           0           0           0           0   \n",
       "2              1           0           0           0           0           0   \n",
       "3              1           0           1           1           0           0   \n",
       "4              1           0           0           0           0           0   \n",
       "5              1           1           0           1           1           1   \n",
       "\n",
       "      1997-07-01  1997-08-01  1997-09-01  1997-10-01  1997-11-01  1997-12-01  \\\n",
       "用户ID                                                                           \n",
       "1              0           0           0           0           0           0   \n",
       "2              0           0           0           0           0           0   \n",
       "3              0           0           0           0           1           0   \n",
       "4              0           1           0           0           0           1   \n",
       "5              1           0           1           0           0           1   \n",
       "\n",
       "      1998-01-01  1998-02-01  1998-03-01  1998-04-01  1998-05-01  1998-06-01  \n",
       "用户ID                                                                          \n",
       "1              0           0           0           0           0           0  \n",
       "2              0           0           0           0           0           0  \n",
       "3              0           0           0           0           1           0  \n",
       "4              0           0           0           0           0           0  \n",
       "5              1           0           0           0           0           0  "
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#将有消费的记为1，没有消费的记为0\n",
    "pivoted_purchase=pivoted_money.applymap(lambda x:1 if x>0 else 0)\n",
    "pivoted_purchase.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "id": "068911E2950947FA8149CAB28C8A12E4",
    "jupyter": {},
    "scrolled": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "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>1997-01-01</th>\n",
       "      <th>1997-02-01</th>\n",
       "      <th>1997-03-01</th>\n",
       "      <th>1997-04-01</th>\n",
       "      <th>1997-05-01</th>\n",
       "      <th>1997-06-01</th>\n",
       "      <th>1997-07-01</th>\n",
       "      <th>1997-08-01</th>\n",
       "      <th>1997-09-01</th>\n",
       "      <th>1997-10-01</th>\n",
       "      <th>1997-11-01</th>\n",
       "      <th>1997-12-01</th>\n",
       "      <th>1998-01-01</th>\n",
       "      <th>1998-02-01</th>\n",
       "      <th>1998-03-01</th>\n",
       "      <th>1998-04-01</th>\n",
       "      <th>1998-05-01</th>\n",
       "      <th>1998-06-01</th>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>用户ID</th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "      <th></th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "      1997-01-01  1997-02-01  1997-03-01  1997-04-01  1997-05-01  1997-06-01  \\\n",
       "用户ID                                                                           \n",
       "1            0.0         NaN         NaN         NaN         NaN         NaN   \n",
       "2            0.0         NaN         NaN         NaN         NaN         NaN   \n",
       "3            0.0         NaN         1.0         0.0         NaN         NaN   \n",
       "4            0.0         NaN         NaN         NaN         NaN         NaN   \n",
       "5            1.0         0.0         NaN         1.0         1.0         1.0   \n",
       "\n",
       "      1997-07-01  1997-08-01  1997-09-01  1997-10-01  1997-11-01  1997-12-01  \\\n",
       "用户ID                                                                           \n",
       "1            NaN         NaN         NaN         NaN         NaN         NaN   \n",
       "2            NaN         NaN         NaN         NaN         NaN         NaN   \n",
       "3            NaN         NaN         NaN         NaN         0.0         NaN   \n",
       "4            NaN         0.0         NaN         NaN         NaN         0.0   \n",
       "5            0.0         NaN         0.0         NaN         NaN         1.0   \n",
       "\n",
       "      1998-01-01  1998-02-01  1998-03-01  1998-04-01  1998-05-01  1998-06-01  \n",
       "用户ID                                                                          \n",
       "1            NaN         NaN         NaN         NaN         NaN         NaN  \n",
       "2            NaN         NaN         NaN         NaN         NaN         NaN  \n",
       "3            NaN         NaN         NaN         NaN         0.0         NaN  \n",
       "4            NaN         NaN         NaN         NaN         NaN         NaN  \n",
       "5            0.0         NaN         NaN         NaN         NaN         NaN  "
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#如果本月进行消费，下月也进行消费，则记为1；如果下月没有消费，则记为0，若本月没有记为消费，则记为nan\n",
    "def purchase_return(data):\n",
    "    status=[]\n",
    "    for i in range(17):#循环17个月\n",
    "        if data[i]==1:#若本月消费\n",
    "            if data[i+1]==1:#下个月也消费\n",
    "                status.append(1)#就记为1\n",
    "            if data[i+1]==0:#下个月不消费，就记为0\n",
    "                status.append(0)\n",
    "        else:\n",
    "            status.append(np.nan)\n",
    "    status.append(np.nan)\n",
    "    return pd.Series(status, index=columns_month)\n",
    "\n",
    "pivoted_purchase_return=pivoted_purchase.apply(purchase_return,axis=1)#axis=1表示计算方向在行的方向上，左右运算\n",
    "pivoted_purchase_return.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "id": "5D22C50A0C864930A8927A84819C742C",
    "jupyter": {},
    "scrolled": true,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<img src=\"https://cdn.kesci.com/rt_upload/5D22C50A0C864930A8927A84819C742C/q6z44vst0t.png\">"
      ],
      "text/plain": [
       "<Figure size 1080x360 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_purchase = (pivoted_purchase_return.sum() / pivoted_purchase_return.count()).reset_index()\n",
    "df_purchase.columns = ['Date', 'PurchaseRate']\n",
    "df_purchase['Date'] = df_purchase.Date.astype(str).apply(lambda x:x[:-3])\n",
    "\n",
    "plt.figure(figsize = (15,5))\n",
    "plt.plot(df_purchase.Date, df_purchase.PurchaseRate)\n",
    "plt.xlabel('时间', fontsize=24)\n",
    "plt.ylabel('回购率', fontsize=24)\n",
    "plt.title('回购率的变化', fontsize=24);"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "892FF6C70BAF44D58A12B376353903E9",
    "jupyter": {},
    "mdEditEnable": false,
    "slideshow": {
     "slide_type": "slide"
    },
    "tags": []
   },
   "source": [
    "* 上图可以看出，在初期用户的回购率并不高，1月的回购率只有15%左右，4月份起回购率稳定在30%左右。\n",
    "* 从每月有回购消费的用户数数据可以看出，回购用户数整体有下降趋势。\n",
    "* 对回购率的分析，再次说明了对于新用户，在其第一次消费后的三个月内是一段重要的时期，需要营销策略积极引导其再次消费及持续消费。\n",
    "* 另外，对于有持续消费的老客，也要适时推出反馈老客户的优惠活动，以加强老客的忠诚度。"
   ]
  }
 ],
 "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": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}