一次线性拟合与决定系数的呈现

 1 import numpy as np
 2 import math
 3 import pandas as pd
 4 import matplotlib.pyplot as plt
 5 import pylab
 6 
 7 df = pd.read_excel(r'XXXX')
 8 
 9 plt.rcParams['font.sans-serif'] = ['SimHei'] #解决中文显示
10 plt.rcParams['axes.unicode_minus'] = False #解决符号无法显示
11 
12 x = df['TDW']
13 y = df['Grain']
14 # 计算相关度
15 def computeCorrelation(x,y):
16     xBar = np.mean(x)
17     yBar = np.mean(y)
18     SSR = 0.0
19     varX = 0.0
20     varY = 0.0
21     for i in range(0,len(x)):
22         diffXXbar = x[i] - xBar
23         difYYbar = y[i] - yBar
24         SSR += (diffXXbar * difYYbar)
25         varX += diffXXbar**2
26         varY += difYYbar**2
27     SST = math.sqrt(varX * varY)
28     return SSR/SST
29 
30 # 计算R平方
31 def polyfit(x,y,degree):
32     results = {}
33     coeffs = np.polyfit(x,y,degree)
34     results['polynomial'] = coeffs.tolist()
35     p = np.poly1d(coeffs)
36     yhat = p(x)
37     ybar = np.sum(y)/len(y)
38     ssreg = np.sum((yhat - ybar)**2)
39     sstot = np.sum((y - ybar)**2)
40     results['determination'] = ssreg/sstot
41     return results
42 
43 result = computeCorrelation(x,y)
44 r = result
45 r_2 = result**2
46 print("r:",r)
47 print("r^2:",r*r)
48 
49 
50 
51 pylab.plot(x,y,'o')
52 
53 z = np.polyfit(x,y,1)
54 p = np.poly1d(z)
55 pylab.plot(x,p(x),'r-')
56 
57 # the line equation:
58 gs = ( 'y = %.4fx + %.4f '%(z[0],z[1]))
59 print(gs)
60 
61 #添加数据标签 还可以优化
62 plt.text(20,9.9,gs,fontsize = 12)
63 plt.text(21,9.5,'$r^{2}=0.343$*',fontsize = 12)
64 plt.title('干物质积累与产量的相关性分析',fontsize=20)
65 plt.xlabel('TDW(t·$ha^{-1}$)',fontsize=12)
66 plt.ylabel('Grain($t·ha^{-1}$)',fontsize=12)
67 plt.show()
68 # plt.savefig('zaodao01.jpg',dpi=600)

最终的呈现效果：

优化版：

 1 import numpy as np
 2 import math
 3 import pandas as pd
 4 import matplotlib.pyplot as plt
 5 import pylab
 6 
 7 plt.rcParams['font.sans-serif'] = 'SimHei' #解决中文显示
 8 plt.rcParams['axes.unicode_minus'] = False #解决符号无法显示
 9 
10 x = df['You']
11 y = df['日产量']
12 
13 xq = x[0:12]
14 xw = x[12:20]
15 xe = x[20:28]
16 yq =y[0:12]
17 yw = y[12:20]
18 ye = y[20:28]
19 
20 #构建总的点 label=''可使其不在图例中显示
21 pylab.plot(x,y,'o',label = '')
22 
23 # 构建三个类型的点
24 pylab.plot(xq,yq,'o', label = '高产')
25 pylab.plot(xw,yw,'o',label = '中产' )
26 pylab.plot(xe,ye,'o', label = '低产')
27 
28 plt.legend(loc='upper left')
29 
30 # 计算相关度
31 def computeCorrelation(x,y):
32     xBar = np.mean(x)
33     yBar = np.mean(y)
34     SSR = 0.0
35     varX = 0.0
36     varY = 0.0
37     for i in range(0,len(x)):
38         diffXXbar = x[i] - xBar
39         difYYbar = y[i] - yBar
40         SSR += (diffXXbar * difYYbar)
41         varX += diffXXbar**2
42         varY += difYYbar**2
43     SST = math.sqrt(varX * varY)
44     return SSR/SST
45 
46 # 计算R平方
47 def polyfit(x,y,degree):
48     results = {}
49     coeffs = np.polyfit(x,y,degree)
50     results['polynomial'] = coeffs.tolist()
51     p = np.poly1d(coeffs)
52     yhat = p(x)
53     ybar = np.sum(y)/len(y)
54     ssreg = np.sum((yhat - ybar)**2)
55     sstot = np.sum((y - ybar)**2)
56     results['determination'] = ssreg/sstot
57     return results
58 
59 result = computeCorrelation(x,y)
60 r = result
61 r_2 = result**2
62 print("r:",r)
63 print("r^2:",r*r)
64 
65 
66 z = np.polyfit(x,y,1)
67 p = np.poly1d(z)
68 pylab.plot(x,p(x),'r-')
69 
70 # the line equation:
71 gs = ( 'y = %.4fx + %.4f '%(z[0],z[1]))
72 print(gs)
73 
74 #添加数据标签
75 plt.text(370,90,gs,fontsize = 12)
76 plt.text(420,86,'$r^{2}=$%.3f*'%r_2,fontsize = 12)
77 plt.title('有效穗数与日产量的相关性分析',fontsize=20)
78 plt.xlabel('有效穗数(10$^{3}$)',fontsize=12)
79 plt.ylabel('日产量(t·ha$^{-1}$)',fontsize=12)
80 #plt.savefig('Rzx03.jpg',dpi=600)

posted on 2019-03-21 13:12 Canvas2018 阅读(448) 评论(0) 收藏举报