{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# 1 Introduction" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This project is about a kind of lottery ticket called \"array 3\", which is a means of raising money by selling numbered tickets and giving prizes to the holders of the number drawn at random. And you have three ways to get the prize:\n", "1. you choose a number from 000 to 999. If your choice matches the prize number，then you get 1040 RMB.\n", "2. you choose one number --A from 0 to 9 twice and choose another different number -- B as the same time,such as'112'. If the prize number includes two 'A'and one 'B'(3 cases in total like 112,121,211)，then you get 346 RMB.\n", "3. you choose three different numbers from 0 to 9, such as'123'.If the prize number includes '1', '2'and '3'(6 cases in total)，then you get 173 RMB." ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import pandas as pd\n", "import numpy as np\n", "import matplotlib.pyplot as plt\n", "import datetime as dt\n", "import scipy.stats as st\n", "%matplotlib inline" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 2 data & clean" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The datas of the first csv come from the [Official website of the lottery ticket](https://www.lottery.gov.cn/historykj/history.jspx?_ltype=pls). I copy the datas and save them as csv." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 2-1 infomation of the data" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "url_all = \"data-all.csv\"\n", "df_all = pd.read_csv(url_all)" ] }, { "cell_type": "code", "execution_count": 3, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/plain": [ "time_serial int64\n", "password object\n", "way1 int64\n", "way1_number int64\n", "way2 int64\n", "way2_number int64\n", "way3 object\n", "way3_number int64\n", "total_sales object\n", "time object\n", "dtype: object" ] }, "execution_count": 3, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_allcopy=df_all.copy()\n", "df_allcopy.dtypes" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The `time_serial` means a serial number to the time.\n", "\n", "The `password` is the prize number\n", "\n", "The `way1`, `way2`, `way3` are the amounts of winning bets for the three ways, and those `way_number` mean the prizes to the ways." ] }, { "cell_type": "code", "execution_count": 4, "metadata": { "scrolled": false }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
time_serialpasswordway1way1_numberway2way2_numberway3way3_numbertotal_salestime
0200984 4 98148104010839346017325,888,2302020/5/27
1200979 3 59422104003463168017325,527,1702020/5/26
2200967 1 811124104003464073417326,909,7862020/5/25
3200954 2 512017104003463470417324,962,7542020/5/24
4200943 2 817743104003465849817326,033,3122020/5/23
\n", "
" ], "text/plain": [ " time_serial password way1 way1_number way2 way2_number way3 \\\n", "0 20098 4 4 9 8148 1040 10839 346 0 \n", "1 20097 9 3 5 9422 1040 0 346 31680 \n", "2 20096 7 1 8 11124 1040 0 346 40734 \n", "3 20095 4 2 5 12017 1040 0 346 34704 \n", "4 20094 3 2 8 17743 1040 0 346 58498 \n", "\n", " way3_number total_sales time \n", "0 173 25,888,230 2020/5/27 \n", "1 173 25,527,170 2020/5/26 \n", "2 173 26,909,786 2020/5/25 \n", "3 173 24,962,754 2020/5/24 \n", "4 173 26,033,312 2020/5/23 " ] }, "execution_count": 4, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_allcopy.head(5)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 2-2 clean the data" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
time_serialpasswordway1way1_numberway2way2_numberway3way3_numbertotal_salespassword_1password_2password_3
time
2020-05-27200984 4 98148104010839346017325888230.0449
2020-05-26200979 3 59422104003463168017325527170.0935
2020-05-25200967 1 811124104003464073417326909786.0718
2020-05-24200954 2 512017104003463470417324962754.0425
2020-05-23200943 2 817743104003465849817326033312.0328
\n", "
" ], "text/plain": [ " time_serial password way1 way1_number way2 way2_number \\\n", "time \n", "2020-05-27 20098 4 4 9 8148 1040 10839 346 \n", "2020-05-26 20097 9 3 5 9422 1040 0 346 \n", "2020-05-25 20096 7 1 8 11124 1040 0 346 \n", "2020-05-24 20095 4 2 5 12017 1040 0 346 \n", "2020-05-23 20094 3 2 8 17743 1040 0 346 \n", "\n", " way3 way3_number total_sales password_1 password_2 \\\n", "time \n", "2020-05-27 0 173 25888230.0 4 4 \n", "2020-05-26 31680 173 25527170.0 9 3 \n", "2020-05-25 40734 173 26909786.0 7 1 \n", "2020-05-24 34704 173 24962754.0 4 2 \n", "2020-05-23 58498 173 26033312.0 3 2 \n", "\n", " password_3 \n", "time \n", "2020-05-27 9 \n", "2020-05-26 5 \n", "2020-05-25 8 \n", "2020-05-24 5 \n", "2020-05-23 8 " ] }, "execution_count": 5, "metadata": {}, "output_type": "execute_result" } ], "source": [ "df_allcopy['time'] = pd.to_datetime(df_allcopy['time'])\n", "\n", "#split the 'password' into three columns\n", "df_allcopy['password_1'] = df_allcopy['password'].str.split(' ').str[0].astype(int)\n", "df_allcopy['password_2'] = df_allcopy['password'].str.split(' ').str[1].astype(int)\n", "df_allcopy['password_3'] = df_allcopy['password'].str.split(' ').str[2].astype(int)\n", "\n", "# chenge the '- -' in df_allcopy['way3'] with '0' and chenge the values into 'int'\n", "df_allcopy.loc[df_allcopy['way3']=='- -', 'way3'] = 0\n", "df_allcopy['way3'] = df_allcopy['way3'].astype(int)\n", "\n", "# dope the '- -' in df_allcopy['total_sales'] and chenge the values into 'float'\n", "df_allcopy.loc[df_allcopy['total_sales']=='- -', 'total_sales'] = np.nan\n", "df_allcopy['total_sales'] = df_allcopy['total_sales'].str.replace(',', '').astype(float)\n", "\n", "\n", "df_A = df_allcopy.set_index('time')\n", "df_A.head(5)" ] }, { "cell_type": "code", "execution_count": 6, "metadata": { "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "DatetimeIndex: 5505 entries, 2020-05-27 to 2004-11-14\n", "Data columns (total 12 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 time_serial 5505 non-null int64 \n", " 1 password 5505 non-null object \n", " 2 way1 5505 non-null int64 \n", " 3 way1_number 5505 non-null int64 \n", " 4 way2 5505 non-null int64 \n", " 5 way2_number 5505 non-null int64 \n", " 6 way3 5505 non-null int32 \n", " 7 way3_number 5505 non-null int64 \n", " 8 total_sales 1345 non-null float64\n", " 9 password_1 5505 non-null int32 \n", " 10 password_2 5505 non-null int32 \n", " 11 password_3 5505 non-null int32 \n", "dtypes: float64(1), int32(4), int64(6), object(1)\n", "memory usage: 473.1+ KB\n" ] } ], "source": [ "df_A.info()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 2-3 Supplementary datas" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The column `total_sales` that means the total sale for one time has only 1345 values, which are not enough.\n", "\n", "The following datas \n", "In the [Official website of the lottery ticket](https://www.lottery.gov.cn/historykj/history.jspx?_ltype=pls), there are links to some new websites, such as [this](https://pdf.sporttery.cn/28200/20107/20107.pdf). And I use a `reptile` to get more datas and save as csv.\n" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 2-3-1 clean the data\n", "I won't show the details of the primeval data by using a data-clean function." ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [], "source": [ "#data-clean function\n", "def csv_clean1(url):\n", " df_new = pd.read_csv(url)\n", " strftime = \"%Y%m%d\" \n", " df_new.loc[:, 'date'] = pd.to_datetime(df_new['date'] , format = strftime)\n", " df_new.loc[:, 'amount'] = df_new['amount'].str.replace(',', '').astype(float)\n", " df_new.loc[:, 'result'] = df_new['result'].str.replace(' ', '').str.replace('、', '').astype(int)\n", " \n", " df1=df_new.set_index('date')\n", " return df1" ] }, { "cell_type": "code", "execution_count": 8, "metadata": { "scrolled": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "DatetimeIndex: 909 entries, 2010-04-28 to 2012-11-05\n", "Data columns (total 2 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 amount 909 non-null float64\n", " 1 result 909 non-null int32 \n", "dtypes: float64(1), int32(1)\n", "memory usage: 17.8 KB\n" ] } ], "source": [ "url1= \"data12897--13500-1.csv\"\n", "url2= \"data13500--13796-1.csv\"\n", "#url3= \"data13796--13806-1.csv\"\n", "df1=pd.concat([csv_clean1(url1), csv_clean1(url2)], axis=0)\n", "df1.info()" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [], "source": [ "#data-clean function\n", "def csv_clean2(url):\n", " df_new = pd.read_csv(url)\n", " \n", " #chenge the time datas in str into datatime\n", " strftime = \"%Y年%m月%d日\" # the time datas of the csv contain Chinese\n", " df_new.loc[:, 'date'] = pd.to_datetime(df_new['date'].str.replace(' ', ''), format= strftime)\n", " \n", " #chenge tother datas in str into str\n", " df_new.loc[:, 'amount'] = df_new['amount'].str.replace(',', '').astype(int)\n", " df_new.loc[df_new['amount']==-1, 'amount'] = np.nan\n", " df_new.loc[:, 'result'] = df_new['result'].str.replace(' ', '').astype(int)\n", " \n", " df1=df_new.set_index('date')\n", " return df1" ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "scrolled": false }, "outputs": [], "source": [ "url4= \"data13806--14445-1--na.csv\"\n", "df_10=csv_clean2(url4)\n", "url5= \"data14000--14233-1.csv\"\n", "url6= \"data14233--14445-1--na.csv\"\n", "url7= \"data14445--15058-1.csv\"\n", "df9=pd.concat([ csv_clean2(url5), csv_clean2(url6), csv_clean2(url7)], axis = 0)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "some datas of `df_10` are same as other datas in `df9`. So, let's merge them." ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "scrolled": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "DatetimeIndex: 2161 entries, 2010-04-28 to 2016-05-09\n", "Data columns (total 2 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 amount 2023 non-null float64\n", " 1 result 2161 non-null float64\n", "dtypes: float64(2)\n", "memory usage: 50.6 KB\n" ] } ], "source": [ "#rename df_10\n", "names={\n", " 'amount': 'amount2',\n", " 'result': 'result2'\n", "}\n", "df_10.rename(columns=names, inplace=True)\n", "\n", "#merge, the datas from different dataframes have different tags\n", "df11 = df9.merge(df_10,left_index=True, right_index= True , how=\"outer\")\n", "\n", "#if the values in df9 is nan ,I replace them with the value in df_10\n", "df11.loc[df11['amount'].isnull(), 'amount']=df11.loc[df11['amount'].isnull(), 'amount2']\n", "df11.loc[df11['result'].isnull(), 'result']=df11.loc[df11['result'].isnull(), 'result2']\n", "\n", "df_B = pd.concat([df1, df11.loc[:, ['amount', 'result']]], axis=0)\n", "df_B.info()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 2-3-2 merge and complete the two dataframe" ] }, { "cell_type": "code", "execution_count": 12, "metadata": { "scrolled": false }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "DatetimeIndex: 5505 entries, 2020-05-27 to 2004-11-14\n", "Data columns (total 16 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 time_serial 5505 non-null int64 \n", " 1 password 5505 non-null object \n", " 2 way1 5505 non-null int64 \n", " 3 way1_number 5505 non-null int64 \n", " 4 way2 5505 non-null int64 \n", " 5 way2_number 5505 non-null int64 \n", " 6 way3 5505 non-null int32 \n", " 7 way3_number 5505 non-null int64 \n", " 8 total_sales 3368 non-null float64\n", " 9 password_1 5505 non-null int32 \n", " 10 password_2 5505 non-null int32 \n", " 11 password_3 5505 non-null int32 \n", " 12 amount 2023 non-null float64\n", " 13 result 2161 non-null float64\n", " 14 total_prize 5505 non-null int64 \n", " 15 ratio_of_prize 3368 non-null float64\n", "dtypes: float64(4), int32(4), int64(7), object(1)\n", "memory usage: 805.1+ KB\n" ] } ], "source": [ "df_total = df_A.merge(df_B,left_index=True, right_index= True , how=\"left\")\n", "df_total.loc[df_total['total_sales'].isnull(),'total_sales' ] = df_total.loc[df_total['total_sales'].isnull(),'amount' ]\n", "\n", "#get the ratio of prize and the total prizes\n", "df_total['total_prize'] = df_total['way1']*df_total['way1_number']+df_total['way2']*df_total['way2_number']+df_total['way3']*df_total['way3_number']\n", "df_total['ratio_of_prize'] = df_total['total_prize']/df_total['total_sales']\n", "df_total.info()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 3 is the password stochastic?" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "if the password is stochastic, each number obeys the distribution (Recording as `P`) independently.\n", "\n", "There are 3 numbers for each period. For convenience, I average them. The mean obeys another distribution(Recording as `Q`) independently." ] }, { "cell_type": "code", "execution_count": 13, "metadata": { "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "4.5 2.75\n" ] } ], "source": [ "#the mean and var of Q\n", "var_theory_Q=np.arange(10).var()/3\n", "mean_theory_Q=np.arange(10).mean()\n", "\n", "print(mean_theory_Q, var_theory_Q)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "when average the mean of the 3 numbers for each period -- Q, we get a distribution(Recording as `NQ`) " ] }, { "cell_type": "code", "execution_count": 14, "metadata": { "scrolled": false }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
meanee_pdf
count27.00000027.00000027.000000
mean4.5642590.5480050.262317
std0.1130610.9641860.125638
min4.363333-1.1654970.047707
25%4.460000-0.3411210.156323
50%4.5700000.5969620.303592
75%4.6608331.3715910.373573
max4.7416672.0609400.394933
\n", "
" ], "text/plain": [ " mean e e_pdf\n", "count 27.000000 27.000000 27.000000\n", "mean 4.564259 0.548005 0.262317\n", "std 0.113061 0.964186 0.125638\n", "min 4.363333 -1.165497 0.047707\n", "25% 4.460000 -0.341121 0.156323\n", "50% 4.570000 0.596962 0.303592\n", "75% 4.660833 1.371591 0.373573\n", "max 4.741667 2.060940 0.394933" ] }, "execution_count": 14, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data_3_1_df = df_total.loc[:, ['password_1', 'password_2', 'password_3']]\n", "#get Q\n", "data_3_1_df['Q'] = data_3_1_df.mean(axis=1)\n", "\n", "#get the mean and var of distributions--NQ\n", "N=200\n", "data_3_1_df['mean'] = data_3_1_df['Q'].rolling(N).mean()\n", "#data_3_1_df['var'] = data_3_1_df['Q'].rolling(N).var()\n", "#data_3_1_df['var_of_mean'] = data_3_1_df['var']/N\n", "\n", "#the mean and var of NQ on theory\n", "mean_theory_NQ= mean_theory_Q\n", "var_theory_NQ= var_theory_Q/N\n", "\n", "#get the independent distributions--NQ\n", "data_3_1_plot = data_3_1_df.iloc[range(N-1, 5505, N ),4:]\n", "\n", "#Standardize NQ\n", "data_3_1_plot['e'] = (data_3_1_plot['mean']-mean_theory_NQ)/(var_theory_NQ)**0.5\n", "data_3_1_plot['e_pdf']=st.norm.pdf(data_3_1_plot['e'])\n", "\n", "data_3_1_plot.describe()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "`N=200`,NQ contains 200 independent distributions--Q, or 600 independent distributions--P.\n", "\n", "According to the `Central limit theorem`, NQ obeys independent normal distributions--`N(mean_theory_NQ,var_theory_NQ)`approximately" ] }, { "cell_type": "code", "execution_count": 15, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "fig, ax = plt.subplots(1)\n", "data_3_1_plot.plot(\n", " ax=ax,\n", " kind='scatter',\n", " x='e',\n", " y='e_pdf',\n", " s=8,\n", " c='black'\n", ");\n", "\n", "\n", "a=np.linspace(-4,4,100)\n", "plt.plot(a,st.norm.pdf(a),color='y')\n", "\n", "\n", "# 5%\n", "x=st.norm.isf(0.025)\n", "y=st.norm.pdf(x)\n", "plt.vlines(st.norm.isf(0.025),0,y,ls='--' ,color='b');\n", "plt.vlines(-st.norm.isf(0.025),0,y,ls='--',color='b');\n", "\n", "# 1%\n", "y2=st.norm.pdf(st.norm.isf(0.005))\n", "plt.vlines(st.norm.isf(0.005),0,y2,ls='--',color='r');\n", "plt.vlines(-st.norm.isf(0.005),0,y2,ls='--',color='r');" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can see that of the 27 observations, two numbers are not within the 95% confidence interval. While, all observations are within the 99% confidence interval.\n", "\n", "But there is still a problem. It seems that most of the observations are on the right side(>0)." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "adjust the `N` to 5500" ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [], "source": [ "data_3_1_df = df_total.loc[:, ['password_1', 'password_2', 'password_3']]\n", "#get Q\n", "data_3_1_df['Q'] = data_3_1_df.mean(axis=1)\n", "\n", "#get the mean and var of distributions--NQ\n", "N=5500\n", "data_3_1_df['mean'] = data_3_1_df['Q'].rolling(N).mean()\n", "\n", "#the mean and var of NQ on theory\n", "mean_theory_NQ= mean_theory_Q\n", "var_theory_NQ= var_theory_Q/N\n", "\n", "#get the independent distributions--NQ\n", "data_3_1_plot = data_3_1_df.iloc[range(N-1, 5505, N ),4:]\n", "\n", "#Standardize NQ\n", "data_3_1_plot['e'] = (data_3_1_plot['mean']-mean_theory_NQ)/(var_theory_NQ)**0.5\n", "data_3_1_plot['e_pdf']=st.norm.pdf(data_3_1_plot['e'])" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "`N=5500`\n", "\n", "According to the `Central limit theorem`, NQ obeys independent normal distributions--`N(mean_theory_NQ,var_theory_NQ)`practically." ] }, { "cell_type": "code", "execution_count": 17, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "fig, ax = plt.subplots(1)\n", "data_3_1_plot.plot(\n", " ax=ax,\n", " kind='scatter',\n", " x='e',\n", " y='e_pdf',\n", " s=8,\n", " c='black'\n", ");\n", "\n", "a=np.linspace(-4,4,100)\n", "plt.plot(a,st.norm.pdf(a),color='y')\n", "\n", "# 5%\n", "x=st.norm.isf(0.025)\n", "y=st.norm.pdf(x)\n", "plt.vlines(st.norm.isf(0.025),0,y,ls='--' ,color='b');\n", "plt.vlines(-st.norm.isf(0.025),0,y,ls='--',color='b');\n", "\n", "# 1%\n", "y2=st.norm.pdf(st.norm.isf(0.005))\n", "plt.vlines(st.norm.isf(0.005),0,y2,ls='--',color='r');\n", "plt.vlines(-st.norm.isf(0.005),0,y2,ls='--',color='r');" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can see that the number is out of the 95% confidence interval, even on the edge the 99% confidence interval. As the same time,the observation is on the right side(>0).\n", "\n", "So, the randomness of the numbers is not as perfect as we think." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "# 4 the sale & ratio" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 4-1 In general" ] }, { "cell_type": "code", "execution_count": 18, "metadata": {}, "outputs": [], "source": [ "#get a dataframe with the complete time series\n", "value = {'nan': np.nan}\n", "time = pd.date_range(start='2004-11-14', end='2020-05-27')\n", "df_time = pd.DataFrame(value, index=time)\n", "a = [ 'way1', 'way2', 'way3', 'total_sales', 'total_prize']\n", "data_4_1_df = df_time.merge(df_total.loc[:, a].dropna(),left_index=True, right_index= True , how=\"left\").iloc[:, 1:]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 4-1-1 the total sales" ] }, { "cell_type": "code", "execution_count": 19, "metadata": { "scrolled": true }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "\n", "DatetimeIndex: 41 entries, 2010-06-30 to 2020-06-30\n", "Freq: BQ-DEC\n", "Data columns (total 7 columns):\n", " # Column Non-Null Count Dtype \n", "--- ------ -------------- ----- \n", " 0 way1 41 non-null float64\n", " 1 way2 41 non-null float64\n", " 2 way3 41 non-null float64\n", " 3 total_sales 41 non-null float64\n", " 4 total_prize 41 non-null float64\n", " 5 ratio_of_prize 41 non-null float64\n", " 6 Y 41 non-null int64 \n", "dtypes: float64(6), int64(1)\n", "memory usage: 2.6 KB\n" ] } ], "source": [ "#get a dataframe with the mean of business quarter\n", "data_4_1_Q=data_4_1_df.resample(\"BQ\").mean()\n", "data_4_1_Q.dropna(inplace=True)\n", "data_4_1_Q['ratio_of_prize'] = data_4_1_Q.loc[:, 'total_prize']/data_4_1_Q.loc[:, 'total_sales']\n", "data_4_1_Q['Y']=data_4_1_Q.index.year\n", "\n", "data_4_1_Q.info()" ] }, { "cell_type": "code", "execution_count": 20, "metadata": { "scrolled": false }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
way1way2way3total_salestotal_prizeratio_of_prizeY
2010-12-318169.4153232431.40322612527.7943552.250945e+071.095191e+070.4865472010
2011-12-318773.9525141963.95810111970.2597772.317832e+071.131766e+070.4882862011
2012-12-317589.4822492025.3698229459.0443791.981185e+079.751048e+060.4921832012
2013-12-316574.7680251757.4294677981.9435741.720629e+078.414256e+060.4890222013
2014-12-315801.3405021585.3405026981.8996421.567297e+077.594769e+060.4845782014
\n", "
" ], "text/plain": [ " way1 way2 way3 total_sales \\\n", "2010-12-31 8169.415323 2431.403226 12527.794355 2.250945e+07 \n", "2011-12-31 8773.952514 1963.958101 11970.259777 2.317832e+07 \n", "2012-12-31 7589.482249 2025.369822 9459.044379 1.981185e+07 \n", "2013-12-31 6574.768025 1757.429467 7981.943574 1.720629e+07 \n", "2014-12-31 5801.340502 1585.340502 6981.899642 1.567297e+07 \n", "\n", " total_prize ratio_of_prize Y \n", "2010-12-31 1.095191e+07 0.486547 2010 \n", "2011-12-31 1.131766e+07 0.488286 2011 \n", "2012-12-31 9.751048e+06 0.492183 2012 \n", "2013-12-31 8.414256e+06 0.489022 2013 \n", "2014-12-31 7.594769e+06 0.484578 2014 " ] }, "execution_count": 20, "metadata": {}, "output_type": "execute_result" } ], "source": [ "#get a dataframe with the mean of year\n", "data_4_1_Y=data_4_1_df.resample(\"Y\").mean()\n", "data_4_1_Y.dropna(inplace=True)\n", "data_4_1_Y['ratio_of_prize'] = data_4_1_Y.loc[:, 'total_prize']/data_4_1_Y.loc[:, 'total_sales']\n", "data_4_1_Y['Y']=data_4_1_Y.index.year\n", "data_4_1_Y.head()" ] }, { "cell_type": "code", "execution_count": 21, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
way1_xway2_xway3_xtotal_sales_xtotal_prize_xratio_of_prize_xYindex_yway1_yway2_yway3_ytotal_sales_ytotal_prize_yratio_of_prize_yQ/Y
index_x
2010-06-307657.4687502062.23437512303.5468752.328062e+071.028595e+070.44182520102010-12-318169.4153232431.40322612527.7943552.250945e+071.095191e+070.4865471.034260
2010-09-308477.0217392239.65217412485.7826092.190322e+071.119144e+070.51094920102010-12-318169.4153232431.40322612527.7943552.250945e+071.095191e+070.4865470.973068
2010-12-318217.9456522879.96739112725.8043482.257920e+071.117566e+070.49495420102010-12-318169.4153232431.40322612527.7943552.250945e+071.095191e+070.4865471.003099
2011-03-317997.9638552315.42168711951.7228922.382693e+071.065117e+070.44702220112011-12-318773.9525141963.95810111970.2597772.317832e+071.131766e+070.4882861.027983
2011-06-309213.1428572253.62637411893.2307692.502334e+071.183722e+070.47304720112011-12-318773.9525141963.95810111970.2597772.317832e+071.131766e+070.4882861.079601
\n", "
" ], "text/plain": [ " way1_x way2_x way3_x total_sales_x \\\n", "index_x \n", "2010-06-30 7657.468750 2062.234375 12303.546875 2.328062e+07 \n", "2010-09-30 8477.021739 2239.652174 12485.782609 2.190322e+07 \n", "2010-12-31 8217.945652 2879.967391 12725.804348 2.257920e+07 \n", "2011-03-31 7997.963855 2315.421687 11951.722892 2.382693e+07 \n", "2011-06-30 9213.142857 2253.626374 11893.230769 2.502334e+07 \n", "\n", " total_prize_x ratio_of_prize_x Y index_y way1_y \\\n", "index_x \n", "2010-06-30 1.028595e+07 0.441825 2010 2010-12-31 8169.415323 \n", "2010-09-30 1.119144e+07 0.510949 2010 2010-12-31 8169.415323 \n", "2010-12-31 1.117566e+07 0.494954 2010 2010-12-31 8169.415323 \n", "2011-03-31 1.065117e+07 0.447022 2011 2011-12-31 8773.952514 \n", "2011-06-30 1.183722e+07 0.473047 2011 2011-12-31 8773.952514 \n", "\n", " way2_y way3_y total_sales_y total_prize_y \\\n", "index_x \n", "2010-06-30 2431.403226 12527.794355 2.250945e+07 1.095191e+07 \n", "2010-09-30 2431.403226 12527.794355 2.250945e+07 1.095191e+07 \n", "2010-12-31 2431.403226 12527.794355 2.250945e+07 1.095191e+07 \n", "2011-03-31 1963.958101 11970.259777 2.317832e+07 1.131766e+07 \n", "2011-06-30 1963.958101 11970.259777 2.317832e+07 1.131766e+07 \n", "\n", " ratio_of_prize_y Q/Y \n", "index_x \n", "2010-06-30 0.486547 1.034260 \n", "2010-09-30 0.486547 0.973068 \n", "2010-12-31 0.486547 1.003099 \n", "2011-03-31 0.488286 1.027983 \n", "2011-06-30 0.488286 1.079601 " ] }, "execution_count": 21, "metadata": {}, "output_type": "execute_result" } ], "source": [ "#merge them\n", "data_4_1_QY=data_4_1_Q.reset_index().merge(data_4_1_Y.reset_index(), on='Y', how='left').set_index('index_x')\n", "data_4_1_QY.head()\n", "data_4_1_QY['Q/Y']= data_4_1_QY['total_sales_x']/data_4_1_QY['total_sales_y']\n", "data_4_1_QY.head()" ] }, { "cell_type": "code", "execution_count": 22, "metadata": { "scrolled": true }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "fig, ax = plt.subplots(1)\n", "\n", "data_4_1_Y['total_sales'].plot(\n", " ax=ax,\n", " color='r',\n", " ylim=[10**7,2.75*10**7]\n", ");\n", "\n", "data_4_1_Q['total_sales'].plot(\n", " ax=ax,\n", " color='y',\n", " ylim=[10**7,2.75*10**7]\n", ");" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "from this,we can see that:\n", "\n", "Sales declined significantly between 2011 and 2016.\n", "\n", "The sales volume rose slowly from 2016 to 2019.(Sales are rising rapidly in 2020, possibly due to missing data.)\n", "\n", "Sales tend to fluctuate between quarters." ] }, { "cell_type": "code", "execution_count": 23, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "fig, ax = plt.subplots(1)\n", "data_4_1_QY.plot(\n", " ax=ax,\n", " y='Q/Y'\n", ");\n", "plt.hlines(1,'2010-06-30','2020-06-30',ls='--' ,color='b');\n", "for i in range(2010,2021,1):\n", " a=f'{i}'+'-06-30'\n", " plt.vlines(a, 0.8,1.1,ls='--',color='r');" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "from this,we can see that:\n", "\n", "Sales fluctuate clearly from quarter to quarter.Sales in the first quarter are often higher than other quarters." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 4-1-2 the ratio of prize\n", "Let's discuss the ratio of prize to sales.\n", "\n", "From 2010 to 2014, this ratio was set at around 50%. \n", "\n", "From August 2014, this ratio was set to 52%." ] }, { "cell_type": "code", "execution_count": 24, "metadata": { "scrolled": true }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "fig, ax = plt.subplots(1)\n", "\n", "data_4_1_Q['ratio_of_prize'].plot(\n", " ax=ax,\n", " color='b'\n", ");\n", "\n", "data_4_1_Y['ratio_of_prize'].plot(\n", " ax=ax,\n", " color='r'\n", ");\n", "\n", "plt.hlines(0.5,'2010-06-30','2014-07-30',ls='--' ,color='b');\n", "plt.hlines(0.52,'2014-07-30','2020-06-30',ls='--' ,color='b');\n", "plt.hlines(0.49,'2010-06-30','2014-07-30',ls='--' ,color='r');\n", "plt.hlines(0.515,'2014-07-30','2020-06-30',ls='--' ,color='r');" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can see that the actual ratio is often smaller than the set ratio. \n", "\n", "Since 2014, the annual average ratio has exceeded the set ratio only once.\n", "\n", "Before 2014, the actual ratio of annual averages was always around 0.45, lower than 0.5." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## 4-2 some details" ] }, { "cell_type": "code", "execution_count": 25, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
way1way2way3total_salestotal_prizeratio_of_prizeyearweekofyeardayofweekdaymonth
2004-11-14NaNNaNNaNNaNNaNNaN20044661411
2004-11-15NaNNaNNaNNaNNaNNaN20044701511
2004-11-16NaNNaNNaNNaNNaNNaN20044711611
2004-11-17NaNNaNNaNNaNNaNNaN20044721711
2004-11-18NaNNaNNaNNaNNaNNaN20044731811
\n", "
" ], "text/plain": [ " way1 way2 way3 total_sales total_prize ratio_of_prize year \\\n", "2004-11-14 NaN NaN NaN NaN NaN NaN 2004 \n", "2004-11-15 NaN NaN NaN NaN NaN NaN 2004 \n", "2004-11-16 NaN NaN NaN NaN NaN NaN 2004 \n", "2004-11-17 NaN NaN NaN NaN NaN NaN 2004 \n", "2004-11-18 NaN NaN NaN NaN NaN NaN 2004 \n", "\n", " weekofyear dayofweek day month \n", "2004-11-14 46 6 14 11 \n", "2004-11-15 47 0 15 11 \n", "2004-11-16 47 1 16 11 \n", "2004-11-17 47 2 17 11 \n", "2004-11-18 47 3 18 11 " ] }, "execution_count": 25, "metadata": {}, "output_type": "execute_result" } ], "source": [ "value = {'nan': np.nan}\n", "time = pd.date_range(start='2004-11-14', end='2020-05-27')\n", "df_time = pd.DataFrame(value, index=time)\n", "a = [ 'way1', 'way2', 'way3', 'total_sales', 'total_prize', 'ratio_of_prize']\n", "data_4_2_df = df_time.merge(df_total.loc[:, a].dropna(),left_index=True, right_index= True , how=\"left\").iloc[:, 1:]\n", "#data_4_2_df['day']=data_4_2_df.index.year()\n", "\n", "data_4_2_df['year']=data_4_2_df.index.year\n", "data_4_2_df['weekofyear']=data_4_2_df.index.weekofyear\n", "data_4_2_df['dayofweek']=data_4_2_df.index.dayofweek\n", "\n", "data_4_2_df['day']=data_4_2_df.index.day\n", "data_4_2_df['month']=data_4_2_df.index.month\n", "\n", "data_4_2_df.head()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 4-2-1 week" ] }, { "cell_type": "code", "execution_count": 26, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
yearweekofyeardayofweektotal_sales
2004-11-142004466NaN
2004-11-152004470NaN
2004-11-162004471NaN
2004-11-172004472NaN
2004-11-182004473NaN
\n", "
" ], "text/plain": [ " year weekofyear dayofweek total_sales\n", "2004-11-14 2004 46 6 NaN\n", "2004-11-15 2004 47 0 NaN\n", "2004-11-16 2004 47 1 NaN\n", "2004-11-17 2004 47 2 NaN\n", "2004-11-18 2004 47 3 NaN" ] }, "execution_count": 26, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data_4_2_1_df=data_4_2_df.loc[:, ['year', 'weekofyear', 'dayofweek', 'total_sales']]\n", "data_4_2_1_df.head()" ] }, { "cell_type": "code", "execution_count": 27, "metadata": { "scrolled": false }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "b=data_4_2_1_df.pivot_table( index=['year','weekofyear'], columns=\"dayofweek\", values='total_sales').dropna()\n", "b.idxmax(axis=1).value_counts().sort_index().plot.bar();" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "It can be clearly seen that the lottery with the results announced on Friday had the best sales in a week. While, the lottery with the results announced on Saturday had the worst sales.\n", "\n", "So people are more inclined to buy lottery tickets with the results announced on Friday." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### 4-2-2 month" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "How about the month?" ] }, { "cell_type": "code", "execution_count": 28, "metadata": { "scrolled": true }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
yearmonthdaytotal_sales
2004-11-1420041114NaN
2004-11-1520041115NaN
2004-11-1620041116NaN
2004-11-1720041117NaN
2004-11-1820041118NaN
\n", "
" ], "text/plain": [ " year month day total_sales\n", "2004-11-14 2004 11 14 NaN\n", "2004-11-15 2004 11 15 NaN\n", "2004-11-16 2004 11 16 NaN\n", "2004-11-17 2004 11 17 NaN\n", "2004-11-18 2004 11 18 NaN" ] }, "execution_count": 28, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data_4_2_2_df=data_4_2_df.loc[:, ['year', 'month', 'day', 'total_sales']]\n", "data_4_2_2_df.head()" ] }, { "cell_type": "code", "execution_count": 29, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "b=indexs_full=data_4_2_2_df.pivot_table(index=['year','month'], columns=\"day\", values='total_sales')\n", "indexs_full=b.isnull().sum(axis=1)<4\n", "\n", "b.loc[indexs_full, :].idxmax(axis=1).value_counts().sort_index().plot.bar();" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "We can't find out much obvious from this picture." ] } ], "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": "3.7.7" } }, "nbformat": 4, "nbformat_minor": 4 }