2014-12-02 95 views
4

我正在運行一個m文件,創建兩個變量ClusWatts_Map。我想將這兩個變量保存到以「.mat」結尾的文件中。這兩個變量的尺寸爲(1152,241,319),其中1152爲360度的經度,增量爲0.3125度,241的緯度爲30S-30N,增量爲0.25度,時間跨度爲319。代碼工作一路才結束,在那裏我得到的錯誤:保存.mat文件,使用-v7.3開關?

[Warning: Variable 'Clus' cannot be saved to a MAT-file whose version is older 
than 7.3. 
To save this variable, use the -v7.3 switch. 
Skipping...] 
[Warning: Variable 'Watts_Map' cannot be saved to a MAT-file whose version is 
older than 7.3. 
To save this variable, use the -v7.3 switch. 
Skipping...] 

我用Matlab版本R2014a,所以會認爲這是最新版本。此外,我已經在較小的空間域上運行了相同的確切代碼(但超過2920個時間步長)而沒有錯誤。

% Clear all variables, initialize counter, indexed by timestep 
clc; 
clear all; 

rain = NaN(1152,241,319); 
Clus = NaN(1152,241,319); 
Areas_Final = NaN(500,319); 
Wattage_Final = NaN(500,319); 
Cluster_size = zeros(319,1); 
Watts_Map = zeros(1152,241,319); 

for year = 2000%:2008; 
    Nyear = sprintf('%04d',year); 

    % Call on the files for each year 
    filename = (['pr_3hr_GFDL-HIRAM-C360_amip_r1i1p1_' num2str(Nyear) '010100-' num2str(Nyear) '123123_subset_TROPICS.nc']); 
    disp(filename) 
    disp(year) 

    rain_rate = ncread(filename,'pr'); 

    % Call on each timestep 
    for i = 960:4:2236; % Approx May 1st-Sep 30th 

     % Set extract subset for region, mask land areas, for each 
     % timestep 

     disp(i) 
     rain = rain_rate(:,:,i); 

     % Eliminate bad/no data points 
     index_rain = (rain >= (5.4e-05)) & (rain < 1e-2); % 0.2mm/hr is min rain rate 

     % Cluster each morning and afternoon matrix 
     Clus(:,:,i) = cluster_it(index_rain); 

     % Calculate cluster areas 
     Areas = cluster_areas_TROPICS(Clus(:,:,i)); 
     Areas_Final(1:length(Areas),i) = Areas; 

     % Calculate cluster wattages 
     Wattage = cluster_wattage(Clus(:,:,i),rain); 

     Cluster_size(i,1) = max(max(Clus(:,:,i))); 

     % Create dummy matricies to populate and use to create the wattage 
     % maps 
     D = zeros(1152,241); 
     E = squeeze(Clus(:,:,i)); 
     for index = 1:Cluster_size(i); 
      D(E == index) = Wattage(index); 
     end 

     Watts_Map(:,:,i) = D; 

     % Clear the dummy matricies 
     clear D E 

    end 

    % Save the output as a .mat file 
    file_out = sprintf(num2str(Nyear), year); 
    matfile = [file_out '_TROPICS_Watts_Maps_inc_Land_low_rain_threshold.mat']; 
    save(matfile, 'Clus', 'Watts_Map'); 

    % Clear unneeded variables to save memory and prevent overwriting 
    clear index_rain rain Areas Wattage Clus file_out filename Areas_Final rain_rate Watts_Map Cluster_size year matfile 

end 

回答

8

即使在當前版本中,默認格式也不是v7.3。你必須將它指定:

save(matfile, '-v7.3', 'var1', 'var2'); 

您還可以設置在 「一般首」 默認:

enter image description here

注意v7.3 does compress data as of R2008a

Compression of -v7.3 MAT-Files

You can store data items that are over 2 gigabytes in size in a MAT-file using the -v7.3 option for the save function. This option was introduced in MATLAB R2006b. With MATLAB R2008a, save now compresses these MAT-files.

我見過的文件大小炸燬v7.3文件,但這是歷史,現在可能更好。

+1

v7.3只是一個略微修改的HDF5格式 - HDF5可以使用壓縮,而快速測試表明MATLAB確實可以壓縮數據。對我來說,保存在v6,v7和v7.3中的數據分別生成大小分別爲14MB,989KB和943KB的文件。這表明v7.3不僅可以壓縮數據,而且有時可以擊敗v7壓縮。至於爲什麼MATLAB在默認情況下不使用v7.3,可能是因爲文件大小對於複雜的嵌套結構/單元陣列有點過分。 – MrAzzaman 2014-12-02 01:32:23

+0

@MrAzzaman很高興知道它壓縮。它不習慣,而且這個事實是一個很大的缺陷。根據[發行說明],它看起來像是R2008a(http://www.mathworks.com/help/matlab/release-notes-older.html#br0xe_f)。 – chappjc 2014-12-02 01:34:06

相關問題