structure is pointlessly nested within itself when I save it. How to unnest?

Maybe kind of a stupid question, but I had a structure saved to disk--let's call it data1.mat. My script would have a few lines like this:

data1 = load('data1.mat');

...

There would be a few of these. Then my script commands would access variables like this:

some_command(data1.some_variable)

I edited some numbers and wanted to update the saved file data1.mat. However, now when I load the updated file, it now has an extra "layer" of "data1". So to access a variable I now need to do:

some_command(data1.data1.some_variable)

This is stupid. How can I undo this?? This should be simple but for some reason I can't figure out how to remove the pointless extra level. I tried selecting all the fields inside the structure layer and saving them to disk, but the same thing happened.

Answer 1

[Steven Lord]

When you save the struct that you received from load after modifying it, use the '-struct' option in your save call.

>> x = 1:10;
>> y = magic(5);
>> S = struct('x', x, 'y', y);
>> save('mymatfile.mat', '-struct', 'S');
>> whos -file mymatfile.mat
  Name      Size            Bytes  Class     Attributes

  x         1x10               80  double              
  y         5x5               200  double              

Answer 2

[Fangjun Jiang]

If you use load() without returning a variable, then the structure won't be created, but then you have to know what variables are in the .mat file. try this:

clear;
a=1;
b=2;
save;
clear
load;
whos
clear
data1=load;
whos

[Alexei M]

Why would you save and load without arguments?

[Alexei M]

Ok I see what you're saying about not returning a variable. But now how do I save it so that the data looks the same as my other .mat files? The other .mat files don't have this extra layer in them.

I would have to transfer EVERY FREAKING FIELD to a workspace variable and then save the ensemble all again, wouldn't I? Is this the only way to do this??

[Alexei M]

Ok I got it.

Loaded it without output variable as you said. Then I looped through the structure and used the

eval 

command to recreate all the fields as variables in the workspace and then selected them all and saved them.

[Fangjun Jiang]

Don't use eval. If your data comes already in a structure, no need to re-create those variables. See Steven's answer.