【数据分析】通过python分析提取matlab工作区中的数据

1. #%%
2. import scipy.io
3. import pandas as pd
4. import numpy as np
6. # Load the .mat file
7. file_path = 'clay_C.mat'
8. mat_data = scipy.io.loadmat(file_path)
10. ## 建立结果表格
11. # 定义行、列名
12. columns = ['0.1', '0.5', '1', '2', '3', '4', '5']
13. index = ['0.1', '0.5', '1', '2', '3', '4', '5']
15. my_psnrs_mean = pd.DataFrame(np.nan,index = index, columns=columns)
17. my_ssims_mean = pd.DataFrame(np.nan,index = index, columns=columns)
19. my_fsims_mean = pd.DataFrame(np.nan,index = index, columns=columns)
21. my_ergas_mean = pd.DataFrame(np.nan,index = index, columns=columns)
22. #%%
23. # 假设MAT文件中包含一个名为'C'的数据变量
24. data = mat_data['C']
25. #%%
26. # 将数据转换为DataFrame
27. df = pd.DataFrame(data)
28. df = df.transpose()
29. #%%
30. for i in range(len(df)):
31.     row = df.iloc[i][0]
32.    
33.     my_df = pd.DataFrame(row)
34.    
35.     psnrs_mean = my_df.iloc[0,0][0][-1]
36.     ssims_mean = my_df.iloc[0,1][0][-1]
37.     fsims_mean = my_df.iloc[0,2][0][-1]
38.     ergas_mean = my_df.iloc[0,3][0][-1]
39.     tau_1 = my_df.iloc[0,4][0][-1]
40.     tau_2 = my_df.iloc[0,5][0][-1]
41.    
42.     my_psnrs_mean.loc[str(tau_1),str(tau_2)] = psnrs_mean
43.    
44.     my_ssims_mean.loc[str(tau_1),str(tau_2)] = ssims_mean
45.    
46.     my_fsims_mean.loc[str(tau_1),str(tau_2)] = fsims_mean
47.    
48.     my_ergas_mean.loc[str(tau_1),str(tau_2)] = ergas_mean
49. #%%
50. my_psnrs_mean*10
51. #%%
52. my_ssims_mean*10
53. #%%
54. my_fsims_mean*10
55. #%%
56. my_ergas_mean*10

复制代码

代码具体分析

这段代码从MAT文件中读取数据，并将其转换为多个Pandas DataFrame，以便进行后续的分析和处置惩罚。以下是代码的具体分析：
1. 导入库和读取MAT文件

1. import scipy.io
2. import pandas as pd
3. import numpy as np
5. # Load the .mat file
6. file_path = 'clay_C.mat'
7. mat_data = scipy.io.loadmat(file_path)

复制代码

这里使用了scipy.io.loadmat函数来读取MAT文件，并将其存储在变量mat_data中。
2. 建立结果表格

1. # 定义行、列名
2. columns = ['0.1', '0.5', '1', '2', '3', '4', '5']
3. index = ['0.1', '0.5', '1', '2', '3', '4', '5']
5. my_psnrs_mean = pd.DataFrame(np.nan,index = index, columns=columns)
6. my_ssims_mean = pd.DataFrame(np.nan,index = index, columns=columns)
7. my_fsims_mean = pd.DataFrame(np.nan,index = index, columns=columns)
8. my_ergas_mean = pd.DataFrame(np.nan,index = index, columns=columns)

复制代码

这里创建了四个空的DataFrame来存储结果，每个DataFrame的行和列都是['0.1', '0.5', '1', '2', '3', '4', '5']。
3. 提取数据并转换为DataFrame

1. # 假设MAT文件中包含一个名为'C'的数据变量
2. data = mat_data['C']
4. # 将数据转换为DataFrame
5. df = pd.DataFrame(data)
6. df = df.transpose()

复制代码

从MAT文件中提取名为'C'的数据变量，将其转换为Pandas DataFrame，并对其进行转置。
4. 迭代每一行并提取指标

1. for i in range(len(df)):
2.     row = df.iloc[i][0]
3.    
4.     my_df = pd.DataFrame(row)
5.    
6.     psnrs_mean = my_df.iloc[0,0][0][-1]
7.     ssims_mean = my_df.iloc[0,1][0][-1]
8.     fsims_mean = my_df.iloc[0,2][0][-1]
9.     ergas_mean = my_df.iloc[0,3][0][-1]
10.     tau_1 = my_df.iloc[0,4][0][-1]
11.     tau_2 = my_df.iloc[0,5][0][-1]
12.    
13.     my_psnrs_mean.loc[str(tau_1),str(tau_2)] = psnrs_mean
14.     my_ssims_mean.loc[str(tau_1),str(tau_2)] = ssims_mean
15.     my_fsims_mean.loc[str(tau_1),str(tau_2)] = fsims_mean
16.     my_ergas_mean.loc[str(tau_1),str(tau_2)] = ergas_mean

复制代码

遍历DataFrame的每一行，并提取每行的各项指标值。具体操作如下：

• 将每一行的第一个元素转换为新的DataFrame my_df。
  
• 提取 psnrs_mean、ssims_mean、fsims_mean、ergas_mean、tau_1 和 tau_2 值。
  
• 将这些值填入相应的结果DataFrame中，使用tau_1和tau_2作为索引。
  

5. 打印结果

1. my_psnrs_mean*10
2. my_ssims_mean*10
3. my_fsims_mean*10
4. my_ergas_mean*10

复制代码

这里将每个结果DataFrame中的值乘以10并显示出来。
总结

该代码的紧张目的是从MAT文件中读取数据，并将此中包罗的各种性能指标提取出来，存储到多个Pandas DataFrame中。每个结果DataFrame的行和列是tau_1和tau_2的组合，存储对应的psnrs_mean、ssims_mean、fsims_mean和ergas_mean值。
原代码

1. %SpaTRPCA_BSDtest200_5method_Corr10
2. clear;clc;close all
3. addpath(genpath('lib'));
4. addpath(genpath('data'));
5. addpath(genpath('competing methods'));
7. rhos = 0.1; %Percent of corrupted pixels
9. % %% Options settings
10. % opts.mu = 1e-2;
11. % opts.tol = 1e-6;
12. % opts.rho = 1.1;
13. % opts.max_iter = 500;0
14. % opts.DEBUG = 0;
15. % opts.tau_F= 2;
16.   
17. method='TCTV';
19. datadir = 'data/CAVE_USED';
20. seqs = dir(datadir);
21. seq3 = seqs(3:end);
22. n_img=length(seq3);
24. temp=zeros(n_img,1);
26. psnrs_TCTV=temp;psnrs_FTCTV=temp;psnrs_HTCTV=temp; psnrs_CTCTV=temp; psnrs_WTCTV=temp; psnrs_GTCTV=temp;
28. ssims_TCTV=temp;ssims_FTCTV=temp;ssims_HTCTV=temp; ssims_CTCTV=temp; ssims_WTCTV=temp; ssims_GTCTV=temp;
30. fsims_TCTV=temp;fsims_FTCTV=temp;fsims_HTCTV=temp; fsims_CTCTV=temp; fsims_WTCTV=temp; fsims_GTCTV=temp;
32. ergas_TCTV=temp;ergas_FTCTV=temp;ergas_HTCTV=temp; ergas_CTCTV=temp; ergas_WTCTV=temp; ergas_GTCTV=temp;
34. sams_TCTV=temp;sams_FTCTV=temp;sams_HTCTV=temp; sams_CTCTV=temp; sams_WTCTV=temp; sams_GTCTV=temp;
36. times_TCTV=temp;times_FTCTV=temp;times_HTCTV=temp; times_CTCTV=temp; times_WTCTV=temp; times_GTCTV=temp;
38. Temp=cell(length(seq3),1);
39. Orig_Multi=Temp;Noisy_Multi=Temp;TCTV_Multi=Temp;FTCTV_Multi=Temp;HTCTV_Multi=Temp;CTCTV_Multi=Temp;WTCTV_Multi=Temp;GTCTV_Multi=Temp;
41. C = cell(1,20);
42. iter=1;
43. sum_iter = 50;
44. sum_time = 0;
45. % for tau_1 = [1,2,3,4,5,6,7,8,9,10]
46. % for tau_1 = [0.1,0.2,0.3,0.4,0.5]
47. for tau_1 = [0.1, 0.5, 1, 2, 3, 4, 5 ]
48. % for tau_2 = [1,2,3,4,5,6,7,8,9,10]
49. % for tau_2 = [0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9,1]
50. for tau_2 = [0.1, 0.5, 1, 2, 3, 4, 5]
51. start_tic = tic;
52. %% Perform recovery
53. for i = 5
54.     % 1 for balloons  2 for beads
55.     % 3 for cd        4 for chart
56.     % 5 for clay      6 for cloth
57.     % 7 for egyptian  8 for feathers
58.     % 9 for flowers   10 for glass
59.     %% load data
60.     fname = seq3(i).name;
61.     pic_TNN_Weight_Lame = [datadir '/' fname];
62.     load(fname)
63.     %% 0-1 Centralization
64.     data_min = min(X(:));
65.     data_max = max(X(:));
66.     X = (X - data_min) / (data_max - data_min);
67.     %% add noise
68. %     [n1,n2,n3] = size(X);
69. %     Xn = X; Yn=reshape(Xn,[n1*n2,n3]);
70. %     ind = find(rand(n1*n2,1)<rhos);
71. %     Yn(ind,:)=rand(length(ind),n3);
72. %     Xn=reshape(Yn,[n1,n2,n3]);
73.     [height, width, band] = size(X);
74.     for ii = 1:band
75.         Xn(:,:,ii) = imnoise(X(:,:,ii), 'salt & pepper', rhos);
76.     end
77.     methods_index = 1;
79.     %% Method 1: TCTV
80. %     disp('......Run TCTV......')
81. %     opts = [];
82. %     opts.rho = 1.25;
83. %     opts.directions = [1,2,3]; % consider the lowrankness and smoothness both along the spatial and spectral directions
84. %     tic
85. %     Xhat_TCTV = TCTV_TRPCA(Xn, opts);
86. %     times_TCTV(i) = toc;
87. %     [psnrs_TCTV(i), ssims_TCTV(i), fsims_TCTV(i), ergas_TCTV(i), sams_TCTV(i)] = HSI_QA(X * 255, Xhat_TCTV* 255);
88. % %     TCTV_psnr = psnr(X , Xhat_TCTV);
89. %     Xhat{methods_index}=Xhat_TCTV;
90. %     div{methods_index}=1-abs((X-Xhat_TCTV));
91. %     methods_index = methods_index+1;
93. %     %% Method 2: FTCTV  Enhanced Fair
94. %     disp('......Run FTCTV......')
95. %     opts = [];
96. %     opts.rho = 1.25;
97. %     opts.directions = [1,2,3]; % consider the lowrankness and smoothness both along the spatial and spectral directions
98. %     opts.tau_F=[tau_1 tau_1, tau_1];
99. %     opts.tauL1=tau_2;
100. %     tic
101. %     Xhat_FTCTV = FTCTV_TRPCA(Xn, opts);
102. %     times_FTCTV(i) = toc;
103. %     [psnrs_FTCTV(i), ssims_FTCTV(i), fsims_FTCTV(i), ergas_FTCTV(i), sams_FTCTV(i)] = HSI_QA(X * 255, Xhat_FTCTV * 255);
104. %     Xhat{methods_index}=Xhat_FTCTV;
105. %     div{methods_index}=1-abs((X-Xhat_FTCTV));
106. %     methods_index = methods_index+1;
107. %
108. %     %% Method 3: HTCTV  Enhanced Huber
109. %     disp('......Run HTCTV......')
110. %     opts = [];
111. %     opts.rho = 1.25;
112. %     opts.directions = [1,2,3]; % consider the lowrankness and smoothness both along the spatial and spectral directions
113. %     opts.tau_F=[tau_1 tau_1, tau_1];
114. %     opts.tauL1=tau_2;
115. %     tic
116. %     Xhat_HTCTV = HTCTV_TRPCA(Xn, opts);
117. %     times_HTCTV(i) = toc;
118. %     [psnrs_HTCTV(i), ssims_HTCTV(i), fsims_HTCTV(i), ergas_HTCTV(i), sams_HTCTV(i)] = HSI_QA(X * 255, Xhat_HTCTV * 255);
119. %     Xhat{methods_index}=Xhat_HTCTV;
120. %     div{methods_index}=1-abs((X-Xhat_HTCTV));
121. %     methods_index = methods_index+1;
122. %
123. %     %% Method 4: CTCTV  Enhanced Cauchy
124. %     disp('......Run CTCTV......')
125. %     opts = [];
126. %     opts.rho = 1.25;
127. %     opts.directions = [1,2,3]; % consider the lowrankness and smoothness both along the spatial and spectral directions
128. %     opts.tau_F=[tau_1 tau_1, tau_1];
129. %     opts.tauL1=tau_2;
130. %     tic
131. %     Xhat_CTCTV = CTCTV_TRPCA(Xn, opts);
132. %     times_CTCTV(i) = toc;
133. %     [psnrs_CTCTV(i), ssims_CTCTV(i), fsims_CTCTV(i), ergas_CTCTV(i), sams_CTCTV(i)] = HSI_QA(X * 255, Xhat_CTCTV * 255);
134. %     Xhat{methods_index}=Xhat_CTCTV;
135. %     div{methods_index}=1-abs((X-Xhat_CTCTV));
136. %     methods_index = methods_index+1;
137. %
138. %     %% Method 5: WTCTV  Enhanced Welsch
139. %     disp('......Run WTCTV......')
140. %     opts = [];
141. %     opts.rho = 1.25;
142. %     opts.directions = [1,2,3]; % consider the lowrankness and smoothness both along the spatial and spectral directions
143. %     opts.tau_F=[tau_1 tau_1, tau_1];
144. %     opts.tauL1=tau_2;
145. %     tic
146. %     Xhat_WTCTV = WTCTV_TRPCA(Xn, opts);
147. %     times_WTCTV(i) = toc;
148. %     [psnrs_WTCTV(i), ssims_WTCTV(i), fsims_WTCTV(i), ergas_WTCTV(i), sams_WTCTV(i)] = HSI_QA(X * 255, Xhat_WTCTV * 255);
149. %     Xhat{methods_index}=Xhat_WTCTV;
150. %     div{methods_index}=1-abs((X-Xhat_WTCTV));
151. %     methods_index = methods_index+1;
153.     %% Method 6: GTCTV  Enhanced General loss
154.     disp('......Run GTCTV......')
155.     opts = [];
156.     opts.rho = 1.25;
157.     opts.directions = [1,2,3]; % consider the lowrankness and smoothness both along the spatial and spectral directions
158.     opts.tau_F=[tau_1 tau_1, tau_1];
159.     opts.tauL1=tau_2;
160.     General_option = 2;
161.     % 2 for Enhanced Geman-McClure loss
162.     % 1 for Enhanced Cauchy loss
163.     tic
164.     Xhat_GTCTV = GTCTV_TRPCA(Xn, opts ,General_option);
165.     times_GTCTV(i) = toc;
166.     [psnrs_GTCTV(i), ssims_GTCTV(i), fsims_GTCTV(i), ergas_GTCTV(i), sams_GTCTV(i)] = HSI_QA(X * 255, Xhat_GTCTV * 255);
167.     Xhat{methods_index}=Xhat_GTCTV;
168.     div{methods_index}=1-abs((X-Xhat_GTCTV));
169.     methods_index = methods_index+1;
171. %     fprintf('\n %d-th image, TCTV   psnr=%.4f, ssim=%.4f, fsim=%.4f\n',i, psnrs_TCTV(i),ssims_TCTV(i),fsims_TCTV(i));
172. %     fprintf('\n              FTCTV  psnr=%.4f, ssim=%.4f, fsim=%.4f\n',   psnrs_FTCTV(i),ssims_FTCTV(i),fsims_FTCTV(i));
173. %     fprintf('\n              HTCTV  psnr=%.4f, ssim=%.4f, fsim=%.4f\n',   psnrs_HTCTV(i),ssims_HTCTV(i),fsims_HTCTV(i));
174. %     fprintf('\n              CTCTV  psnr=%.4f, ssim=%.4f, fsim=%.4f\n',   psnrs_CTCTV(i),ssims_CTCTV(i),fsims_CTCTV(i));
175. %     fprintf('\n              WTCTV  psnr=%.4f, ssim=%.4f, fsim=%.4f\n',   psnrs_WTCTV(i),ssims_WTCTV(i),fsims_WTCTV(i));
176. %     fprintf('\n              GTCTV  psnr=%.4f, ssim=%.4f, fsim=%.4f\n',   psnrs_GTCTV(i),ssims_GTCTV(i),fsims_GTCTV(i));
177.    
178.     psnrs_i=[psnrs_TCTV(i),psnrs_FTCTV(i),psnrs_HTCTV(i),psnrs_CTCTV(i),psnrs_WTCTV(i),psnrs_GTCTV(i)];
179.     ssims_i=[ssims_TCTV(i),ssims_FTCTV(i),ssims_HTCTV(i),ssims_CTCTV(i),ssims_WTCTV(i),ssims_GTCTV(i)];  
180.     ergas_i=[ergas_TCTV(i),ergas_FTCTV(i),ergas_HTCTV(i),ergas_CTCTV(i),ergas_WTCTV(i),ergas_GTCTV(i)];
181.    
182.     disp(['PSNR_i=       ',num2str(psnrs_i)])  
183.     disp(['SSIM_i=       ',num2str(ssims_i)])
184.     disp(['ERGAS_i=       ',num2str(ergas_i)])
185.    
186. %     Orig_Multi{i}=X;Noisy_Multi{i}=Xn;TCTV_Multi{i}=Xhat_TCTV;FTCTV_Multi{i}=Xhat_FTCTV;HTCTV_Multi{i}=Xhat_HTCTV;
187. %     j=1;
188. end
189. methodName={'TCTV','FTCTV','HTCTV','CTCTV','WTCTV','GTCTV'};
190. enList=[1,2,3,4,5,6];
191. psnrs_all=[psnrs_TCTV,psnrs_FTCTV,psnrs_HTCTV,psnrs_CTCTV,psnrs_WTCTV,psnrs_GTCTV];
192. psnrs_mean=mean(psnrs_all,1);
193. ssims_all=[ssims_TCTV,ssims_FTCTV,ssims_HTCTV,ssims_CTCTV,ssims_WTCTV,ssims_GTCTV];
194. ssims_mean=mean(ssims_all,1);
195. fsims_all=[fsims_TCTV,fsims_FTCTV,fsims_HTCTV,fsims_CTCTV,fsims_WTCTV,fsims_GTCTV];
196. fsims_mean=mean(fsims_all,1);
197. ergas_all=[ergas_TCTV,ergas_FTCTV,ergas_HTCTV,ergas_CTCTV,ergas_WTCTV,ergas_GTCTV];
198. ergas_mean=mean(ergas_all,1);
199. times_all=[times_TCTV,times_FTCTV,times_HTCTV,times_CTCTV,times_WTCTV,times_GTCTV];
200. times_mean=mean(times_all,1);
202. C{iter} = {psnrs_mean, ssims_mean, fsims_mean, ergas_mean, tau_1, tau_2};
203. epoch_time = toc(start_tic);
204. sum_time = sum_time+epoch_time;
205. fprintf('Progress:%d/%d. Epoch Time:%.2f. Progress Time:%.2f. Time needs:%.2f.\n',iter,sum_iter,epoch_time/60,sum_time/60,((sum_time)/iter)*(sum_iter-iter)/60);
206. iter = iter+1;
207. end
208. end
210. %% Show results
211. showHSIResult(Xhat,X,Xn,div,methodName,enList,31,31);

复制代码

免责声明：如果侵犯了您的权益，请联系站长，我们会及时删除侵权内容，谢谢合作！更多信息从访问主页：qidao123.com:ToB企服之家，中国第一个企服评测及商务社交产业平台。