How to dynamically change name of table?

This code below is just a very simple example of some analysis I am doing with my experimental results.

I am running 5 cycles of an experiment, results come out in a long txt file, I want to plot everything with a different color based on cycle number. So I am creating sub-tables from the main table based on the parameter of interest. This simple code works parfectly. However, it's fine because it is only 5 runs. What if it was 100?

Is there a more efficient way to assign table names dynamically? I am new with Matlab and I am using my real life needs to practice and learn.

Thank you!

T=readtable("Output-C1_Cu.txt","NumHeaderLines",2);

NT1=T(T.Var2==1,:);
NT2=T(T.Var2==2,:);
NT3=T(T.Var2==3,:);
NT4=T(T.Var2==4,:);
NT5=T(T.Var2==5,:);

plot(NT1,"Var4","Var6",LineWidth=1.5, Color=[0 0.4470 0.7410])
hold on
plot(NT2,"Var4","Var6",LineWidth=1.5, Color=[0.8500 0.3250 0.0980])
hold on
plot(NT3,"Var4","Var6",LineWidth=1.5, Color=[0.9290 0.6940 0.1250])
hold on
plot(NT4,"Var4","Var6",LineWidth=1.5, Color=[0.4940 0.1840 0.5560])
hold on
plot(NT5,"Var4","Var6",LineWidth=1.5, Color=[0.4660 0.6740 0.1880])
hold off

xlim([0,1.5])
xlabel('V cell (V)','FontWeight','bold')
ylabel("dQ/dV cell (mAh/V)","FontWeight","bold")
title('Treated Cu - OG Si')
legend({'Cycle 1','Cycle 2','Cycle 3','Cycle 4','Cycle 5',}, ...
    'Location','southeast')

[Stephen23]

"Is there a more efficient way to assign table names dynamically? "

No, there is not. Every way is slow, inefficient, complex, and pointlessly obfuscates code:

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

MATLAB is designed around arrays and indexing. Indexing is exactly how you would handle "What if it was 100?", because indexing trivially expands to any size (hint: try using a loop). Indexing is exactly how experienced MATLAB users would do it, or by using the inbuilt tools for processing groups of data:

https://www.mathworks.com/help/matlab/matlab_prog/perform-calculations-by-group-in-table.html

https://www.mathworks.com/help/matlab/matlab_prog/grouped-calculations-in-tables-and-timetables.html

Your approach of copy-pasting lots of code or dynamically naming variables ... is best avoided.

"I am new with Matlab and I am using my real life needs to practice and learn."

Then forget about dynamic variable names. That is a path you really don't want to go down:

https://www.mathworks.com/matlabcentral/discussions/tips/847916-experts-of-matlab-how-did-you-learn-any-advice-for-beginner-intermediate-users/2556966#reply_2556966

[Marina]

Thank you for all the links!

Answer 1

[Voss]

One way to plot the table data for each cycle separately, without having to create a new table variable for each cycle, is to use findgroups and splitapply.

% make up a table
Var = randi([1,10],40,6);
Var(:,4) = Var(:,4)/10;
T = array2table(Var)

T = 40x6 table
    Var1    Var2    Var3    Var4    Var5    Var6
    ____    ____    ____    ____    ____    ____

      7      10      3      0.1       6       9 
      2       5      9      0.4       8       1 
      9       8      8      0.6       8       8 
      8       9      6      0.6       9       7 
      7       3      1      0.4       2       8 
      5       8      7      0.5       4       2 
      8       9      7        1       3       5 
      1       9      9      0.9       5       2 
      7       4      8      0.6       4       2 
      7       4      2      0.9       1       6 
     10       9      6      0.2       5       5 
      8       1      2      0.9       9       2 
      8       2      9      0.6       6       8 
     10       6      7      0.9       1       1 
     10       7      2      0.9       3       4 
      5       5      6      0.3      10      10 

% plot Var6 vs Var4 separately for each Var2 group
figure
hold on
[G,GID] = findgroups(T.Var2);
F = @(varargin)plot(varargin{4},varargin{6},LineWidth=1.5);
splitapply(F,T,G)
hold off

xlim([0,1.5])
xlabel('V cell (V)','FontWeight','bold')
ylabel("dQ/dV cell (mAh/V)","FontWeight","bold")
title('Treated Cu - OG Si')
legend("Cycle "+GID,'Location','southeast')

Also running your code to plot the first 5 groups, for comparison/validation:

NT1=T(T.Var2==1,:);
NT2=T(T.Var2==2,:);
NT3=T(T.Var2==3,:);
NT4=T(T.Var2==4,:);
NT5=T(T.Var2==5,:);

figure
hold on
plot(NT1,"Var4","Var6",LineWidth=1.5, Color=[0 0.4470 0.7410])
plot(NT2,"Var4","Var6",LineWidth=1.5, Color=[0.8500 0.3250 0.0980])
plot(NT3,"Var4","Var6",LineWidth=1.5, Color=[0.9290 0.6940 0.1250])
plot(NT4,"Var4","Var6",LineWidth=1.5, Color=[0.4940 0.1840 0.5560])
plot(NT5,"Var4","Var6",LineWidth=1.5, Color=[0.4660 0.6740 0.1880])
hold off

xlim([0,1.5])
xlabel('V cell (V)','FontWeight','bold')
ylabel("dQ/dV cell (mAh/V)","FontWeight","bold")
title('Treated Cu - OG Si')
legend({'Cycle 1','Cycle 2','Cycle 3','Cycle 4','Cycle 5',}, ...
    'Location','southeast')

[Marina]

Thank you for providing a proposed direction and functions for me to look up!

[Voss]

You're welcome! Any questions, let me know. If it turns out to be an acceptable solution to the problem, please "Accept" this answer. Thanks!

[Marina]

It did work, thank you! I will go through each line of your code to understand what you did and learn from it.

[Voss]

You're welcome! Let me know if you have any questions.

[Marina]

Would you mind commenting on the syntax of this line?

F = @(varargin)plot(varargin{4},varargin{6},LineWidth=1.5)

I have an idea about it, but it is better to know exactly how everything works, so I can apply in other other cases.

[Voss]

That creates an anonymous function F which takes some arguments varargin and plots the 6th argument against the 4th argument with a line width of 1.5.

The next line

splitapply(F,T,G)

runs that function F on the table T for each group stored in G (groups were determined based on T.Var2 using findgroups). splitapply on a table passes each column of the table as an input to the function you specify (in this case F), so that's why the 6th argument and 4th argument are used in F to plot the 6th column of the table vs the 4th, for each group.

[Marina]

Great, thank you.

[Voss]

You're welcome!

[Voss]

By the way, it may be convenient to collect each groups' data together, into say a cell array of tables, rather than just plotting it directly. Then you can plot it afterward or do whatever else you need to do on a group-by-group basis. Here's one way of doing that:

% make up a table
Var = randi([1,10],40,6);
Var(:,4) = Var(:,4)/10;
T = array2table(Var)

T = 40x6 table
    Var1    Var2    Var3    Var4    Var5    Var6
    ____    ____    ____    ____    ____    ____

      8       8       4       1      10       5 
      8       5      10     0.2       5       6 
      8       3       6     0.8       7       6 
      7       7       3     0.3      10       3 
     10       9       6     0.9      10       1 
      2       9       8     0.5       6       7 
      2      10       2     0.1       4       6 
      5       7       3     0.9       7       3 
      4       8       4     0.7       8       4 
      7       6       5     0.1      10       5 
      7       6       7     0.3       7      10 
      8       5       1     0.6       6      10 
      6       6       7     0.5       4       5 
      4       4       1     0.1       8       2 
      8       1       5     0.7       5       3 
      7       6       1     0.8       9       8 

% collect groups of data
[G,GID] = findgroups(T.Var2);
F = @(varargin){table(varargin{:})};
C = splitapply(F,T,G)

C = 10x1 cell array
    {4x6 table}
    {3x6 table}
    {6x6 table}
    {2x6 table}
    {6x6 table}
    {6x6 table}
    {3x6 table}
    {4x6 table}
    {4x6 table}
    {2x6 table}

C{1} % group 1 data

ans = 4x6 table
    Var1    Var2    Var3    Var4    Var5    Var6
    ____    ____    ____    ____    ____    ____

      8      1        5     0.7      5       3  
      2      1        2     0.6      8       7  
      2      1       10     0.8      9       7  
     10      1        6     0.3      4       7  

% plot Var6 vs Var4 separately for each Var2 group
figure
hold on
for ii = 1:numel(C)
    plot(C{ii}.Var4,C{ii}.Var6,LineWidth=1.5)
end
hold off

xlim([0,1.5])
xlabel('V cell (V)','FontWeight','bold')
ylabel("dQ/dV cell (mAh/V)","FontWeight","bold")
title('Treated Cu - OG Si')
legend("Cycle "+GID,'Location','southeast')

Here the anonymous function F

F = @(varargin){table(varargin{:})};

makes a cell array containing one table of data. Then (as before) splitapply runs that function for each group, so you get C, a cell array containing a table of data for each group.