How can I link data from separate arrays?

Suppose you have an array [4 7 9 4 8 3] which represents a list of parameter values within a database. For example ‘4’ is the thickness of a plate which has an associated array containing location and temperature values (5x2 double) [0.55 0.16; 0.60 0.16; 0.65 0.15; 0.7 0.14; 0.75 0.12]. Each of the thickness values has an associated 5x2 array which is unique.

So the question I have is this: Is there a way I can link the location and temperature data arrays with the thickness values so that I can plot all the arrays which correspond to thickness value of ‘4’ without doing this manually?

[dpb]

Insufficient information -- we know nothing of where these other data are nor how they're defined.

The answer to the general question is "of course" but it does have the caveat that you have to have some piece of information somewhere to use to be able to create that link--whether it's a file-naming convention, the sequence of the associated data in a cell array, ...

[Matthew Weltevreden]

“you have to have some piece of information somewhere to use to be able to create that link” could you perhaps give an example or link for this please?

[Stephen23]

"...an example or link for this please?"

https://www.mathworks.com/help/matlab/matlab_prog/array-comparison-with-relational-operators.html

https://www.mathworks.com/company/newsletters/articles/matrix-indexing-in-matlab.html

[dpb]

Be more helpful to us to have information on how the corollary data are stored; generated; etc., ... to have some idea on how they are linked/associated.

[Matthew Weltevreden]

field1 = 'Data1';
field2 = 'Data2';
field3 = 'Data3';

Thickness1 = 19.0;
Thickness2 = 22.0;
Thickness3 = 32.0;

Temp_data1 = [0,-0.05;0.2,-0.13;0.4,0.25;0.6,0.64;0.8,0.56;1,-0.063];
Temp_data2 = [0,0.36;0.2,-0.15;0.4,-0.1;0.6,0.21;0.8,0.47;1,0.35];
Temp_data3 = [0,0.64;0.2,-0.22;0.4,-0.42;0.6,-0.11;0.8,0.28;1,-0.027];

value1={Thickness1,Temp_data1};
value2={Thickness2,Temp_data2};
value3={Thickness3,Temp_data3};

database = struct(field1,value1,field2,value2,field3,value3);

input_thickness = 20;

I've provided some more information to help with the question, assuming the data is part of a struct i would like to extract all the temp data which is larger than a given input thickness in this case 20 (i.e. Temp_data2 & Temp_data3)

Answer 1

[dpb]

Making harder than needs must be --

Thickness=[19.0;22.0;32.0];
Temp_data=cat(3l [0,-0.05;0.2,-0.13;0.4,0.25;0.6,0.64;0.8,0.56;1,-0.063];
                 [0,0.36;0.2,-0.15;0.4,-0.1;0.6,0.21;0.8,0.47;1,0.35];
                 [0,0.64;0.2,-0.22;0.4,-0.42;0.6,-0.11;0.8,0.28;1,-0.027]);

input_thickness = 20;
isWanted=(Thickness>input_thickness);   % logical vector of locations wanted
data=Temp_data(:,:,isWanted);           % return temperature data wanted

[Matthew Weltevreden]

very helpful, thank you!

[Steven Lord]

This approach has Thickness and Temp_data in separate variables. If Matthew Weltevreden wants to store the data in a combined variable, one question to ask is which form of "slicing" should be easier.

Thickness=[19.0;22.0;32.0];
Temp_data=cat(3, [0,-0.05;0.2,-0.13;0.4,0.25;0.6,0.64;0.8,0.56;1,-0.063], ...
                 [0,0.36;0.2,-0.15;0.4,-0.1;0.6,0.21;0.8,0.47;1,0.35], ...
                 [0,0.64;0.2,-0.22;0.4,-0.42;0.6,-0.11;0.8,0.28;1,-0.027]);

structContainingArrays = struct('Thickness', Thickness, 'Temperature', Temp_data);

arrayContainingStructs = struct('Thickness', Thickness(1), ...
    'Temperature', Temp_data(:, :, 1));
arrayContainingStructs(2) = struct('Thickness', Thickness(2), ...
    'Temperature', Temp_data(:, :, 2));
arrayContainingStructs(3) = struct('Thickness', Thickness(3), ...
    'Temperature', Temp_data(:, :, 3));
whos

  Name                        Size             Bytes  Class     Attributes

  Temp_data                   6x2x3              288  double              
  Thickness                   3x1                 24  double              
  arrayContainingStructs      1x3               1064  struct              
  structContainingArrays      1x1                648  struct              

Yes, I know I probably could have written those last lines more simply.

structContainingArrays makes it easier to "slice" the data looking at all the Thicknesses or all the Temperatures at once. It's more difficult to use that to get both Thickness and Temperature for one of the parameter values.

structContainingArrays.Thickness

ans = 3×1

19 22 32

arrayContainingStructs slices the data differently. Getting the Thickness and Temperature for one of the parameter values is easier while getting all the Thickness values or all the Temperatures is harder.

arrayContainingStructs(2)

ans = struct with fields:
      Thickness: 22
    Temperature: [6×2 double]

Another way to store this data may be to create a table array.

struct2table(arrayContainingStructs)

ans = 3x2 table
    Thickness    Temperature 
    _________    ____________

       19        {6×2 double}
       22        {6×2 double}
       32        {6×2 double}

[dpb]

"Another way to store this data may be to create a table array."

Agreed, and probably what I would do in real life...

[Matthew Weltevreden]

Thanks for the advice. I was also wondering what could be done if the Temp_data arrays are not identical i.e. if a 8x2 double was also part of the same database such as below?

[0.2,-0.23;0.25,0.029;0.3,0.3;0.35,0.56;0.4,0.78;0.45,0.93;0.5,1.021;0.55,1.02]

[dpb]

That would then require something like the table with the 2D arrays a cells; you wouldn't be able to use the 3D array "trick" as the planes have to be the same dimension(*) to do that.

(*) It would require augmenting to the maximum size with NaN or somesuch to do that.

Answer 2

[Hernia Baby]

The code below is one I assumed from your infomation.

Therefore, it might be different from what you want to do.

As the other members said, you need to give us enough input data.

Thickness = [4 7 9 4 8 3];

% create datas by the number of thickness datas
name = ("parts"+char('A'+(0:length(Thickness)-1'))');
LocTemp(:,:,1) = [0.55 0.16; 0.60 0.16; 0.65 0.15; 0.7 0.14; 0.75 0.12];
for i = 1:length(Thickness)-1
  LocTemp(:,:,i+1) = LocTemp(:,:,i) + 0.01;
end

% seperate data into location and temperature
Location = squeeze(LocTemp(:,1,:))';
Temperature =  squeeze(LocTemp(:,2,:))';

% integration as cell type
for i = 1:length(Thickness)
    C(i,:) = {name(i),Thickness(i),Location(i,:),Temperature(i,:)};
end

% legends
leg = {'name' 'thickness' 'location' 'temperature'};

% convert type
S = cell2struct(C, leg, 2);

% output example
S(1)

>> S(1)

name: "partsA"

thickness: 4

location: [0.5500 0.6000 0.6500 0.7000 0.7500]

temperature: [0.1600 0.1600 0.1500 0.1400 0.1200]

[Matthew Weltevreden]

also works very well with previous example! Thank you!