How to name and save the results using the name of imported file ?

Hello everyone,

I have many files containing data. Each file has a different name based on the date for example (data_2020_03_12_08_30_03.mat).

I want to perform my calculations and save my results in a variable with the same name as the imported file for example (Tempreture_data_2020_03_12_08_30_03.mat).

just you to know I do not want to import all data files at once. I am doing this manualy. I want to import only one file every time but I want the results save in a diffrent name every time depend on the name of imported file.

Could you help me how to name and save the results using the name of imported file??

[Stephen23]

"I want to perform my calculations and save my results in a variable with the same name as the imported file for example (Tempreture_data_2020-03-12_08.30.03.mat)."

Do NOT do that: https://www.mathworks.com/matlabcentral/discussions/tips/849901-tutorial-why-variables-should-not-be-named-dynamically-eval

Your filename includes dot/period characters and minus characters: are those valid characters in MATLAB variable names? No, they are not. What do you expect MATLAB to do when you try to magically create invalid variable names? So already your anti-pattern approach is showing some signs of why it should be avoided.

[Ahmad Al-Issa]

thank you for explanation

[Ahmad Al-Issa]

yes you are right. I changed the file name to (data_2020_03_12_08_30_03). there is typos.

Answer 1

[Stephen23]

A much better approach is to store the data in e.g. a structure array:

P = 'absolute or relative path to where the files are saved';
S = dir(fullfile(P,'*.mat'));
for k = 1:numel(S)
    F = fullfile(S(k).folder,S(k).name);
    S(k).data = load(F);
end

All of your imported data are contained in the structure S, for example the 2nd file:

S(2).name % filename
S(2).data % imported data (as a scalar structure)

If all of the files contain the same variables then you can easily concatenate them into one non-scalar structure with all of those variables as fields:

D = [S.data] % optional

That makes your data easy to access:

https://www.mathworks.com/help/matlab/matlab_prog/access-multiple-elements-of-a-nonscalar-struct-array.html

You did not tell us anything about the content of those MAT files (e.g. how many variables they contain, what their names are), so this advice remains quite general.

[Ahmad Al-Issa]

Thanks for replay, but I do not know how to implement your code.

just to know i do not want to import all data files at once. I am doing this manualy. i want to import only one file every time but i want the results save in a diffrent name every time depend on the name of imported file.

[Stephen23]

" I am doing this manualy. i want to import only one file every time but i want the results save in a diffrent name every time depend on the name of imported file."

Okay, so just import the data and change its name using the variable viewer. Good luck.

[Ahmad Al-Issa]

"Okay, so just import the data and change its name using the variable viewer. Good luck."

How can I do that by code? I want one code for all files.

[Ahmad Al-Issa]

I said I would import the data manually but that doesn't mean doing everything manually. I have 300 files, so it's not practical to change the name every time. I need code to save the results with a different name based on the file name.

Each file contains many variables. But this is not important. I just want to know how to save results based on file name to apply this code to all data files without duplicating saved files.

[Stephen23]

"How can I do that by code? I want one code for all files."

That contradicts your statement "I am doing this manualy".

So you want a piece of code that magically defines different variable names each time you LOAD a MAT file. Well, apart from the fact that this is very bad data design, it is very hard to give advice because you have given absolutely no information on the content of those MAT files. Perhaps they already contain variables of different names, in which case your inadvisable goal would be very easily achieved. Perhaps each MAT file contains one variable, perhaps many variables. I have no idea, because you have not given any information.

[Stephen23]

"I said I would import the data manually but that doesn't mean doing everything manually. I have 300 files, so it's not practical to change the name every time. I need code to save the results with a different name based on the file name."

The code I gave in my answer works for any number of files.

What is your specific reason that you cannot use the normal, efficient, standard approach to importing multiple files (which is to use indexing) ? This is the approach shown in the MATLAB documentation:

https://www.mathworks.com/help/matlab/import_export/process-a-sequence-of-files.html

Question: how are you going to process all of those files once you have them in the MATLAB workspace? With 300 files I doubt that you will "manually" process each one separately, in which case I guesss you imagine using some kind of loop... in which case you really need to read this:

https://www.mathworks.com/matlabcentral/discussions/tips/849901-tutorial-why-variables-should-not-be-named-dynamically-eval

"Save" mean exporting data from an application to a file. But you seem to mean something different.

"Each file contains many variables. But this is not important."

It is important.

"I just want to know how to save results based on file name to apply this code to all data files without duplicating saved files."

Simple you use a loop and indexing. My answer shows you an easy way to do this. For some reason you seem to want to make this much harder for yourself. It would be interesting to know the specific reason why.

In any case, the real question is: how are you planning on processing all of this data?

[Ahmad Al-Issa]

I have more than 300 data file. each file has measurement data about temperature, flow rate, etc. the only different thing is the name of file which based on the date. i can reach to the date during the file

data.fileinfo.starttime = 2020_03_12_08_30_03

I need a code that apply to each file individually but save the result in different name every time.

[Stephen23]

"I need a code that apply to each file individually but save the result in different name every time."

Sure, it is clear what you are asking.

However in general your approach should be avoided. It will make processing your data slow, complex and inefficient. That is why I asked how you are going to process this data. Because that information really will define what are efficient and simple approaches for storing the imported data when you import it into MATLAB.

I strongly doubt that you "need" to do that: https://www.mathworks.com/matlabcentral/answers/105936-how-to-make-dynamic-variable-names-a1-a2-an-with-for-loop-using-eval-num2str#answer_332454

[Ahmad Al-Issa]

The reason is simple because the data files are not all available at the same time. Every day I get new reading files and therefore I need a code to perform the results based on the name of these files.

[Ahmad Al-Issa]

this is the codel

% Load measurement data

[file,path] = uigetfile('*.mat', 'Select an *.mat-file');

data = importdata(strcat(path,file));

% for example the tempreture

T1 = data.S7_Medien1.x82_U01_FT1HD_Hyd__ltankTemperaturIstwert.data

% so I want to save T1 in diffrent name based on the file name (date) like T1_2020_03_12_08_30_03

Is this clear or not ?

[Stephen23]

"The answer is simple because the data files are not all available at the same time."

Okay, then there is nothing stopping you from using a non-scalar structure, much like the one in my answer. No magical variable names required. Import each MAT file, store the data using indexing:

• if you know the total number of files in advance then you could preallocate the entire structure when you start,
• otherwise start with an empty structure and enlarge it for each file.

My guess is that you might prefer to use dynamic fieldnames:

https://www.mathworks.com/help/matlab/matlab_prog/generate-field-names-from-variables.html

[Ahmad Al-Issa]

% Load measurement data

[file,path] = uigetfile('*.mat', 'Select an *.mat-file');

data = importdata(strcat(path,file));

% for example the tempreture

T1 = data.S7_Medien1.x82_U01_FT1HD_Hyd__ltankTemperaturIstwert.data

% so I want to save T1 in diffrent name based on the file name (date) like T1_2020_03_12_08_30_03

[Stephen23]

"Is this clear or not ?"

Sure, it is very clear. You have repeated it several times.

However my advice that this should be avoided is apparently not clear enough. Is this clear or not?:

YOUR APPROACH SHOULD BE AVOIDED.

What exactly is stopping you from using simple and efficient indexing?

[Stephen23]

Some notes on your code:

• Do not use PATH as a variable name.
• LOAD is preferred for MAT files. IMPORTDATA is best avoided.
• Prefer using FULLFILE rather than STRCAT on filepaths.

Try something like this. Before you import any files define a counter and an empty structure:

N = 0;
S = struct('name',{},'data',{});

Then for each file that you "manually" import:

[F,P] = uigetfile('*.mat', 'Select a *.mat-file');
D = load(fullfile(P,F));
D.filename = fullfile(P,F); % optional (but useful if you concatenate the data together)
N = N+1;
S(N).data = D;
S(N).name = fullfile(P,F);

All of the imported file data will be stored in the structure S. You can trivially access all of the filenames and imported filedata using basic indexing. For example, for the 2nd file:

S(2).name % the filename (which can include minus and dot characters !)
S(2).data % the imported file data (a scalar structure)

Note how by design this is more robust than your approach. If all of the MAT files contain the same variables, you can easily create one non-scalar structure of all of the imported data:

D = [S.data]

[Ahmad Al-Issa]

thank you, sir,