Python数据挖掘实用案例——自动售货机销售数据分析与应用（二）

groupby2 = data.groupby(by='月份', as_index=False).agg({'设备编号': f, '订单编号': f})
groupby2.columns = ['月份', '设备数量', '订单数量']
# 绘制图形
plt.figure(figsize=(10, 4))
plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False
fig, ax1 = plt.subplots()  # 使用subplots函数创建窗口
ax1.plot(groupby2['月份'], groupby2['设备数量'], '--')
ax1.set_yticks(range(0, 350, 50))  # 设置y1轴的刻度范围
ax1.legend(('设备数量',), loc='upper left', fontsize=10)
ax2 = ax1.twinx()  # 创建第二个坐标轴
ax2.plot(groupby2['月份'], groupby2['订单数量'])
ax2.set_yticks(range(0, 100000, 10000))  # 设置y2轴的刻度范围
ax2.legend(('订单数量',), loc='upper right', fontsize=10)
ax1.set_xlabel('月份')
ax1.set_ylabel('设备数量（台）')
ax2.set_ylabel('订单数量（单）')
plt.title('订单数量和自动售货机数量之间的关系')
plt.show()
gruop3 = data.groupby(by='市', as_index=False).agg({'总金额（元）':sum, '设备编号':f})
gruop3['销售总额'] = np.round(gruop3['总金额（元）'], 2)
gruop3['平均销售总额'] = np.round(gruop3['销售总额'] / gruop3['设备编号'], 2)
plt.bar(gruop3['市'].values.tolist(), gruop3['平均销售总额'].values.tolist(), color='#483D8B')
# 添加数据标注
for x, y in enumerate(gruop3['平均销售总额'].values):
    plt.text(x - 0.4, y + 100, '%s' %y, fontsize=8)
plt.xlabel('城市')
plt.ylabel('平均销售总额（元）')
plt.title('各市自动售货机平均销售总额')
plt.show()

# 销售额前10的商品
group4 = data.groupby(by='商品ID', as_index=False)['总金额（元）'].sum()
group4.sort_values(by='总金额（元）', ascending=False, inplace=True)
d = group4.iloc[: 10]
x_data = d['商品ID'].values.tolist()
y_data = np.round(d['总金额（元）'].values, 2).tolist()
bar = (Bar(init_opts=opts.InitOpts(width='800px',height='600px'))
       .add_xaxis(x_data)
       .add_yaxis('', y_data, label_opts=opts.LabelOpts(font_size=15))
       .set_global_opts(title_opts=opts.TitleOpts(title='畅销前10的商品'),
                        yaxis_opts=opts.AxisOpts(axislabel_opts=opts.LabelOpts(
                            formatter='{value}',font_size=15)),
                        xaxis_opts=opts.AxisOpts(type_='category',
                            axislabel_opts=opts.LabelOpts({'interval': '0'}, font_size=15, rotate=30))))
bar.render_notebook()
h = group4.iloc[-10: ]
x_data = h['商品ID'].values.tolist()
y_data = np.round(h['总金额（元）'].values, 2).tolist()
bar = (Bar()
       .add_xaxis(x_data)
       .add_yaxis('', y_data, label_opts=opts.LabelOpts(position='right'))
       .set_global_opts(title_opts=opts.TitleOpts(
                        title='滞销前10的商品'),
                        xaxis_opts=opts.AxisOpts(
                            axislabel_opts={'interval': '0'}))
       .reversal_axis()
       )
grid = Grid(init_opts=opts.InitOpts(width='600px', height='400px'))
grid.add(bar, grid_opts=opts.GridOpts(pos_left='18%'))
grid.render_notebook()

group5 = data.groupby(by=['市', '设备编号'], as_index=False)['总金额（元）'].sum()
group5.sort_values(by='总金额（元）', ascending=False, inplace=True)
b = group5[: 10]
label = []
# 销售额前10的设备及其所在市
for i in range(len(b)):
    a = b.iloc[i, 0] + str(b.iloc[i, 1])
    label.append(a)
x = np.round(b['总金额（元）'], 2).values.tolist()
y = range(10)
plt.bar(x=0, bottom=y, height=0.4, width=x, orientation='horizontal')
plt.xticks(range(0, 80000, 10000))  # 设置x轴的刻度范围
plt.yticks(range(10), label)
for y, x in enumerate(np.round(b['总金额（元）'], 2).values):
    plt.text(x + 500, y - 0.2, "%s" %x)
plt.xlabel('总金额（元）')
plt.title('销售额前10的设备及其所在市')
plt.show()
l = group5[-10: ]
label1 = []
for i in range(len(l)):
    a = l.iloc[i, 0] + str(l.iloc[i, 1])
    label1.append(a)
x = np.round(l['总金额（元）'], 2).values.tolist()
y = range(10)
plt.bar(x=0, bottom=y, height=0.4, width=x, orientation='horizontal')
plt.xticks(range(0, 4, 1))  # 设置x轴的刻度范围
plt.yticks(range(10), label1)
for y, x in enumerate(np.round(l['总金额（元）'], 2).values):
    plt.text(x, y, "%s" %x)
plt.xlabel('总金额（元）')
plt.title('销售额后10的设备及其所在市')
plt.show()

l_b = group5[group5['总金额（元）'] < 100]
lb = l_b.groupby(by='市', as_index=False)['设备编号'].count()
x_data = lb['市'].values.tolist()
y_data = lb['设备编号'].values.tolist()
bar = (Bar(init_opts=opts.InitOpts(width='500px', height='400px'))
       .add_xaxis(x_data)
       .add_yaxis('', y_data)
       .set_global_opts(title_opts=opts.TitleOpts(
                title='各市销售额小于100的设备数量'))
       )
bar.render_notebook()

group6 = data.groupby(by='支付状态')['支付状态'].count()
method = group6.index.tolist()
num = group6.values.tolist()
pie_data = [(i, j) for i, j in zip(method, num)]
pie = (Pie()
     .add('', pie_data, label_opts=opts.LabelOpts(formatter='{b}:{c}({d}%)'))
     .set_global_opts(title_opts=opts.TitleOpts(title='订单支付方式占比')))
pie.render_notebook()

group7 = data.groupby(by='下单时间段')['购买用户'].count()
times = group7.index.tolist()
num = group7.values.tolist()
pie_data_2 = [(i, j) for i, j in zip(times, num)]
pie = (Pie()
       .add('', pie_data_2, label_opts=opts.LabelOpts(formatter='{b}:{c}({d}%)'),
            radius=[60, 200], rosetype='radius', is_clockwise=False)
       .set_global_opts(title_opts=opts.TitleOpts(title='各消费时段的订单用户占比'))
       )
pie.render_notebook()

import pandas as pd
import matplotlib.pyplot as plt
from statsmodels.tsa.stattools import adfuller as ADF
from statsmodels.graphics.tsaplots import plot_acf
from statsmodels.stats.diagnostic import acorr_ljungbox
from statsmodels.tsa.arima_model import ARIMA
data = pd.read_csv('../tmp/order.csv', encoding='gbk')