using eval with save function

Question

Ham Man 2022-7-11

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1756925-using-eval-with-save-function

评论： Image Analyst 2022-7-11

采纳的回答： the cyclist

在 MATLAB Online 中打开

I'm using eval function to save a .mat file and I'm getting error:

I appreciate any hints!

try_num = 1;
pathdatasave = (['E:\abc\try' num2str(try_num) '\']); 
eval(['save([pathdatasave loc_vel_thresh' num2str(thresh) '_pln' num2str(pln_num) '_try' num2str(try_num) ...
     '_between two planes x1=' num2str(Xl) '-x2=' num2str(Xr)...
     '_myfile.mat],x_end_pln'  num2str(pln_num) '_try' num2str(try_num) ');' ])
Error using vertcat
Dimensions of arrays being concatenated are not consistent.

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Stephen23 2022-7-11

编辑：Stephen23 2022-7-11

在 MATLAB Online 中打开

"I appreciate any hints!"

Don't use EVAL to call SAVE.

Don't force meta-data into variable names.

The variable name that you are trying to create

x_end_pln' num2str(pln_num) '_try' num2str(try_num)

tells us that you made the mistake of forcing meta-data into your variable names (i.e. pln_num and try_num), which when you try to access dynamically forces you into writing slow, complex, inefficient, obfuscated, insecure, buggy code that is hard to debug:

https://www.mathworks.com/matlabcentral/answers/304528-tutorial-why-variables-should-not-be-named-dynamically-eval

In short, bad data design leads to bad code, which is what you are finding out now.

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

the cyclist 2022-7-11

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1756925-using-eval-with-save-function#answer_1004175

在 MATLAB Online 中打开

It can be tricky to construct the string needed, from a combination of text and numerics. I would recommend a few things here.

The most important one is to not use eval. Instead, use sprintf to create the required strings for the filename and/or any variables you want to save, and then put those strings into the save command itself.

Second, my advice would be to always display the string itself to the command line, so that you can see what it looks like. In your case, it looks like this:

% I defined a bunch of these variables arbitrarily, because you had not
% defined them in the code you posted
try_num = 1;
thresh = 2;
pln_num = 3;
Xl = 4;
Xr = 5;
pathdatasave = (['E:\abc\try' num2str(try_num) '\']); 
savestring = ['save([pathdatasave loc_vel_thresh' num2str(thresh) '_pln' num2str(pln_num) '_try' num2str(try_num) ...
     '_between two planes x1=' num2str(Xl) '-x2=' num2str(Xr)...
     '_myfile.mat],x_end_pln'  num2str(pln_num) '_try' num2str(try_num) ');' ]
savestring = 'save([pathdatasave loc_vel_thresh2_pln3_try1_between two planes x1=4-x2=5_myfile.mat],x_end_pln3_try1);'

Notice how the displayed string is using the literal string 'pathdatasave' instead of the actual string you wanted to swap in? I think that is the main technical issue.

So, using a simpler example case, I would do something like

try_num = 1;
pln_num = 2;
filename = sprintf('E:\\abc\\try%d\\',try_num)
filename = 'E:\abc\try1\'
varname  = sprintf('blah_pln%d',pln_num)
varname = 'blah_pln2'

and then just save directly without using eval:

save(filename,varname)

3 个评论
显示 1更早的评论隐藏 1更早的评论

Stephen23 2022-7-11

编辑：Stephen23 2022-7-11

在 MATLAB Online 中打开

"Isn't this due to the single quotation?"

No, it is due to the fact that you are attempting to generate an invalid variable name.

The single quotes are not part of the character vector, they are just an artifact of how it is displayed.

Once we actually create the very badly-named variable blah_pln2 the code works correctly:

varname = 'blah_pln2';
blah_pln2 = 3;
save('test.mat',varname)
whos -file test.mat
  Name           Size            Bytes  Class     Attributes

  blah_pln2      1x1                 8  double              

But clearly, having numbered variable names is a sign that you are doing something wrong.

the cyclist 2022-7-11

What @Stephen23 says is accurate. Probably the first thing I should have posted is this tutorial about why dynamic variable names are usually a terrible idea.

请先登录，再进行评论。

Answer 2

Image Analyst 2022-7-11

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1756925-using-eval-with-save-function#answer_1004165

在 MATLAB Online 中打开

Don't use eval. See the FAQ:

https://matlab.fandom.com/wiki/FAQ#Why_is_it_advised_to_avoid_using_the_%22eval%22_function?

It's simply not necessary.

try_num = 1;
% Construct the folder name.
folder = fullfile('E:\abc\try', num2str(try_num)); 
if ~isfolder(folder)
    mkdir(folder);
end
% Construct the base file name.
baseFileName = sprintf('loc_vel_thresh%d_pln%d_try%d_between two planes x1=%d-x2=%d_myfile.mat',...
    thresh, pln_num, Xl, Xr, pln_num)
% Construct the full name.
fullFileName = fullfile(folder, baseFileName)
% Save the mat file.
save(fullFilename, 'x_end_pln', 'pln_num', 'try_num');

4 个评论
显示 2更早的评论隐藏 2更早的评论

Ham Man 2022-7-11

Thank you so much every one

Image Analyst 2022-7-11

So did this sprintf & save solution work for you? If so, could you click the "Accept this answer" link? Thanks in advance. 🙂 Otherwise tell us what's not working yet.

请先登录，再进行评论。