I have some code that keeps failing to open and read two csv files out of four. Can someone explain what is wrong with either the code or the csv files

I have a piece of code that plots force extension, stress strain and then the average stress strain of four csv files. I have used this multiple times on similar data sets but for some reason this code doesn't like all of the samples this time.

Here is the code

My m. file is called post process

%%
%3M Conventional-Consolodation Two-Stage-Sintering
%
%%
clear
close all

load('D:\Axiom\3M_CS\OXmcdata.mat')
%
Nsamples = 4;
%
L = 90; % coupon gauge section (mm)
A = [17.9 18.75 16.62 13.59]; % coupon cross-sectional area (mm^2)
%
figure(1);
hold;
figure(2);
hold
figure(3);
hold
%
for k = 1:Nsamples
    %
    file_name = ['Sample' num2str(k) '.csv'];
    %
    T = readtable(file_name,'VariableNamingRule','preserve');
    %
    [test(k).disp,test(k).force,test(k).eyy] = post_fun(T);
    %
    figure(1)
    h(k) = plot(test(k).disp,test(k).force/1000);
    set(h(k),'LineWidth',2);
    leg_string{k} = ['Sample ' num2str(k)];
    %
    figure(3)
    g(k) = plot(test(k).eyy*100,test(k).force/A(k));
    set(g(k),'LineWidth',2);
    %
    leg_string{k} = ['Sample ' num2str(k)];
end
%
eyymax = max(test(1).eyy);
eyymin = min(test(1).eyy);
for k = 2:Nsamples
    %
    auxmax = max(test(k).eyy);
    auxmin = min(test(k).eyy);
    %
    eyymax = min([eyymax auxmax]);
    eyymin = max([eyymin auxmin]);
end    
eyysim = linspace(eyymin,eyymax,21)';
%
figure(2)
for k = 1:Nsamples
    test(k).stress = test(k).force/A(k);
    P(k,1:2) = polyfit(test(k).eyy*100,test(k).stress,1);
    stresslin(:,k) = polyval(P(k,:),eyysim*100);
end
meanstress = mean(stresslin,2);
stdstress = std(stresslin,0,2);
%
%     
% g(k) = plot(eyysim*100,);
% set(g(k),'LineWidth',2);
% %
%
figure(1)
set(gcf,'Position',[200 200 1024 768],'Color',[1 1 1]);
set(gca,'FontSize',24);
xlabel('Displacement (mm)','FontSize',32,'Interpreter','latex');
ylabel('force (kN)','FontSize',32,'Interpreter','latex');
box on
grid on
legend(leg_string,'FontSize',28,'Location','northwest','Interpreter','latex');
%
%
figure(2)
set(gcf,'Position',[200 200 1024 768],'Color',[1 1 1]);
set(gca,'FontSize',24);
xlabel('Strain (\%)','FontSize',32,'Interpreter','latex');
ylabel('Stress (MPa)','FontSize',32,'Interpreter','latex');
plot(eyysim*100,meanstress,'k','LineWidth',2)
plot(eyysim*100,meanstress+stdstress,'k--','LineWidth',2)
plot(eyysim*100,meanstress-stdstress,'k--','LineWidth',2)
plot(OXmcdata(:,1),OXmcdata(:,2),'DisplayName','Farhandi et al (2021)','LineWidth',3,'Color','r','LineStyle','--');
title({'Graph showing the mean stress/strain behaviour of' ... 
    '3M spread tow manufactured using consolodation at $200^\circ C$'... 
     'for 2 hours and sintering at $1150^\circ C$ for four hours'},'Interpreter','latex')
box on
grid on
%legend(leg_string,'FontSize',28,'Location','northwest','Interpreter','latex');
%
figure(3)
set(gcf,'Position',[200 200 1024 768],'Color',[1 1 1]);
set(gca,'FontSize',24);
xlabel('Strain (\%)','FontSize',32,'Interpreter','latex');
ylabel('Stress (MPa)','FontSize',32,'Interpreter','latex');
plot(OXmcdata(:,1),OXmcdata(:,2),'DisplayName','Farhandi et al (2021)','LineWidth',3,'Color','r','LineStyle','--');
box on
grid on
legend(leg_string,'FontSize',28,'Interpreter','latex');
%
function [disp,force,eyy] = post_fun(T)
%
disp = table2array(T(:,4));
force = table2array(T(:,end));
eyy = table2array(T(:,10));
%
index = isnan(disp) | isnan(force) | isnan(eyy) | disp < 0 | force < 0 | eyy < 0;
%
disp(index) = [];
force(index) = [];
eyy(index) = [];
%
[~,index] = max(force);
disp(index+1:end) = [];
force(index+1:end) = [];
eyy(index+1:end) = [];
%
end

Please hel me figure out what is wrong.

% Unable to perform assignment because the left and right sides have a different number of elements.
% 
% Error in post_process (line 31)
%     h(k) = plot(test(k).disp,test(k).force/1000);

This is the error

I don't really understand Matlab errors so it is quite difficult to figure out.

Thanks

Alex

Answer 1

[Walter Roberson]

Replace

    h(k) = plot(test(k).disp,test(k).force/1000);
    set(h(k),'LineWidth',2);

with

    h{k} = plot(test(k).disp,test(k).force/1000, 'LineWidth', 2);

and replace

    g(k) = plot(test(k).eyy*100,test(k).force/A(k));
    set(g(k),'LineWidth',2);

with

    g{k} = plot(test(k).eyy*100,test(k).force/A(k), 'LineWidth', 2);

Answer 2

[Voss]

When processing Sample3.csv, post_fun removes every row of table T because index, defined here

index = isnan(disp) | isnan(force) | isnan(eyy) | disp < 0 | force < 0 | eyy < 0;

is all true. That is, each row of T has a disp, force, and/or eyy value that is NaN or negative.

Here's running the relevant code just for that file to show that test(k) contains empty disp, force, and eyy fields.

k = 3;
file_name = ['Sample' num2str(k) '.csv'];
T = readtable(file_name,'VariableNamingRule','preserve');
[test(k).disp,test(k).force,test(k).eyy] = post_fun(T);


function [disp,force,eyy] = post_fun(T)
%
disp = table2array(T(:,4));
force = table2array(T(:,end));
eyy = table2array(T(:,10));
%
index = isnan(disp) | isnan(force) | isnan(eyy) | disp < 0 | force < 0 | eyy < 0;
%
disp(index) = [];
force(index) = [];
eyy(index) = [];
%
[~,index] = max(force);
disp(index+1:end) = [];
force(index+1:end) = [];
eyy(index+1:end) = [];
%
end


% disp, force, and eyy are all empty vectors (0-by-1):
disp(test(k))

     disp: [0x1 double]
    force: [0x1 double]
      eyy: [0x1 double]

% relevant columns of table T for reference:
disp(T(:,[4 end 10]))

    Y-displacement (Strain Gauge 1)    ForceADC1Channel1    Eyy - RC (Strain Gauge 1)
    _______________________________    _________________    _________________________

                      NaN                       NaN                       NaN        
                        0                    -436.1                  2.22e-16        
              -0.00012479                   -425.83                  1.43e-05        
                0.0012639                   -419.38                 -4.62e-05        
              -0.00013617                   -405.34                 -3.35e-06        
              -0.00050161                    -416.8                  1.03e-05        
                0.0011603                   -445.61                  3.41e-06        
              -0.00063076                   -392.68                  5.07e-05        
               -0.0017988                    -306.6                  4.01e-05        
               -0.0069887                    -201.1                0.00018296        
                -0.011494                   -38.675                0.00024613        
                -0.019983                    171.35                0.00035093        
                -0.027295                    338.02                0.00044636        
                -0.034253                    560.21                0.00057036        
                -0.042832                    767.67                0.00068995        
                -0.051284                    973.48                0.00080948        
                -0.060255                    1191.9                0.00094657        
                -0.068178                    1369.5                 0.0010625        
                -0.076328                    1547.9                  0.001154        
                -0.083529                    1746.9                 0.0012349        
                -0.090759                    1911.9                 0.0013251        
                 -0.10058                      2106                 0.0014168        
                 -0.10835                    2281.5                 0.0015124        
                 -0.11624                    2440.9                 0.0015719        
                 -0.12422                    2622.9                 0.0017376        
                 -0.13314                    2763.6                 0.0018018        
                 -0.14089                    2912.6                 0.0018513        
                 -0.14957                    3067.6                 0.0019891        
                  -0.1579                    3188.6                  0.002053        
                 -0.16572                    3318.4                 0.0020814        
                 -0.17349                    3417.8                 0.0020724        
                 -0.18165                    3534.2                 0.0021593        
                    -0.19                    3640.3                 0.0022839        
                 -0.19748                    3739.1                 0.0023523        
                 -0.20667                    3833.9                 0.0023563        
                  -0.2183                    3932.5                 0.0024687        
                  -0.2241                    4003.9                  0.002477        
                 -0.23367                    4076.8                 0.0024916        
                 -0.24154                    4142.2                 0.0025482        
                 -0.25153                    4211.4                 0.0025995        
                 -0.26116                    4283.4                  0.002648        
                 -0.26761                    4327.6                 0.0026228        
                 -0.30327                    3962.7                 0.0023672        
                 -0.44024                    1284.9                0.00097759        
                 -0.44791                    1454.3                 0.0010561        
                 -0.45414                    1652.2                 0.0011356        
                 -0.45803                    1733.9                 0.0012132        
                    -0.46                      1728                 0.0011809        
                 -0.45971                    1721.9                 0.0012358        
                 -0.45913                    1716.1                 0.0011932        

When you plot some empty vectors you get zero lines:

temp = plot(test(k).disp,test(k).force/1000)

temp = 
  0x1 empty Line array.

Which cannot be stored in h(k), since h(k) is a single element of h, i.e., enough to store exactly one line object.

h(k) = plot(test(k).disp,test(k).force/1000);

% ^^^^ h(k) is a slot for exactly one line, but

% ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ this plot call returns zero lines

That's the reason for the error.

It's clear the code was written expecting that plot call to return a single line object, which is an assumption that is violated in this case. How should the situation that the plotted vectors are empty be handled? Maybe post_fun shouldn't be removing rows of T that contain negative disp, force, or eyy?

[A Poyser]

It is strange because I have used this code multiple times and all of the csv files come from the same place are formatted the same and all work. As far as I can see sample 1 and 2 are the same as 3 and 4 in terms of the formatting. The first line is always NaN. The code has worked pretty well before, which is why I am so confused by the error.

[Voss]

As I showed, the error is because in Sample3.csv, each row contains a NaN or negative disp, force, and/or eyy, so all rows are filtered out before plotting. Thus the plot command returns zero line objects instead of one line object, which cannot be stored as h(k).

[A Poyser]

Both sample 1 and sample 2 have rows containing a NaN or negative disp, force, and/or eyy as well.

[Voss]

But not all rows, in those files.

[A Poyser]

got it, so the complete lack of a number is what is messing it up?

[Voss]

For Sample3.csv, because post_fun has removed all the rows due to each row having a NaN or negative, test(k).disp, test(k).force, and test(k).eyy are all empty vectors (0-by-1):

disp(test(k))
     disp: [0x1 double]
    force: [0x1 double]
      eyy: [0x1 double]

When you plot some empty vectors you get zero lines:

temp = plot(test(k).disp,test(k).force/1000)
temp = 
  0x1 empty Line array.

Which cannot be stored in h(k), since h(k) is a single element of h, i.e., enough to store exactly one line object. That's what the error message is trying to say.