191106_爬数据绘图

这个单子是一个姑娘，最开始觉得很简单，就是一些数据的可视化，结果有个数据爬取，要用固定的方式爬取，就难在那个地方了，还有学到了如何绘制一图多表

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
%matplotlib inline

test1=pd.read_excel('discharge.xlsx')
test1=test1.drop(['Unnamed: 0'],axis=1)
test1.head()

	year	discharge
0	2017-01-01	185.0
1	2017-01-02	185.0
2	2017-01-03	185.0
3	2017-01-04	185.0
4	2017-01-05	190.0

test2=pd.read_excel('temperature.xlsx')
test2.head()

	year	0axt
0	2017-01-01	28
1	2017-01-02	27
2	2017-01-03	27
3	2017-01-04	29
4	2017-01-05	40

all_test=pd.merge(test1,test2)

all_test.head()

	year	discharge	0axt
0	2017-01-01	185.0	28
1	2017-01-02	185.0	27
2	2017-01-03	185.0	27
3	2017-01-04	185.0	29
4	2017-01-05	190.0	40

all_test['year'] = all_test['year'].apply(lambda x: pd.Timestamp(x))

all_test.columns

Index(['year', 'discharge', '0axt'], dtype='object')

all_test['年']=all_test['year'].apply(lambda x: x.year)

all_test.head()

	year	discharge	0axt	年
0	2017-01-01	185.0	28	2017
1	2017-01-02	185.0	27	2017
2	2017-01-03	185.0	27	2017
3	2017-01-04	185.0	29	2017
4	2017-01-05	190.0	40	2017

test_2017=all_test.loc[(all_test['年']==2017)]

test_2018=all_test.loc[(all_test['年']==2018)]

# set up plot
plt.figure(figsize=(10,5))
plt.xlabel('day of year 2017')
# plot data
plt.plot(test_2017.index,test_2017['0axt'],'r',label='temperature')
# plt.plot(data['year'],data['snowprop']*100,'b',label='snow cover %')
# plt.plot(data['year'],100-data['snowprop']*100,'c',label='snow free cover %')
plt.plot(test_2017.index,test_2017['discharge']/100.,'g',label='river flow / 100')
plt.legend(loc='best')

# set up plot
plt.figure(figsize=(10,5))
plt.xlabel('day of year 2018')
# plot data
plt.plot(test_2018.index,test_2018['0axt'],'r',label='temperature')
# plt.plot(data['year'],data['snowprop']*100,'b',label='snow cover %')
# plt.plot(data['year'],100-data['snowprop']*100,'c',label='snow free cover %')
plt.plot(test_2018.index,test_2018['discharge']/100.,'g',label='river flow / 100')
plt.legend(loc='best')

a=test_2017.to_numpy
np.save('2016.npy',a)

b=test_2018.to_numpy
np.save('2017.npy',b)