How to create legend for two error bar plots

The only problem with my code is, that the legend line gives error messages and doesn't show up in the graph. I want a legend with 4 elements (The data of graph 1, the fit of graph 1, the data of graph 2, the fit of graph 2). I don't want the error bars to show in the legend. My code:

function [fitresult, gof, goft] = Linearized_plot_both(inverse_square_radius,time,er,inverse_square_radiust,timet,ert)
%%Fit: 'Linearized Fit'.
[xData, yData] = prepareCurveData( inverse_square_radius, time );
[xDatat, yDatat] = prepareCurveData( inverse_square_radiust, timet );

% Set up fittype and options.
ft = fittype( 'poly1' );

% Fit model to data.
[fitresult, gof] = fit( xData, yData, ft );
[fitresultt, goft] = fit( xDatat, yDatat, ft );

% Plot fit with data.
figure( 'Name', 'Comparing Linearized Fit' );
x= linspace(0,11e5);
y=(fitresult.p1).*x + (fitresult.p2);
h=plot(xData,yData,'.k',x,y,'b');
h(1).MarkerSize = 12;
h(2).LineWidth = 1;
hold on;
axis([0 10.5e5 0 41.5]);
grid on;
hEB=errorbar(xData,yData,er,'.k', 'MarkerFaceColor','k','MarkerEdgeColor','k','MarkerSize',12);

xt= linspace(0,11e5);
yt=(fitresultt.p1).*x + (fitresultt.p2);
ht=plot(xDatat,yDatat,'.k',xt,yt,'r');
ht(1).MarkerSize = 12;
ht(2).LineWidth = 1;
hold on;
hEBt=errorbar(xDatat,yDatat,ert,'.k', 'MarkerFaceColor','k','MarkerEdgeColor','k','MarkerSize',12);

legend([h,ht],'time vs. inverse_square_radius (Data)','Linearized Fit (Data)','time vs. inverse_square_radius (Theory)','Linearized Fit (Theory)');

% Label axes
xlabel inverse_square_radius
ylabel time

The produced graph:

Way I want my legend to look (just for both graphs):

%

[dpb]

I don't see anything obvious but can't run your code nor duplicate the intended effect without a fair effort. How about posting a simple test case with included data that illustrates the issue; perhaps even in doing that you'll uncover the problem...

[Nicholas Pfaff]

My Data:

inverse_square_radius =
 1.0e+06 *

    1.0000
    0.4444
    0.2500
    0.1600
    0.1111
    0.0816
    0.0625
    0.0494
    0.0400
    0.0331
    0.0278
    0.0237
    0.0204
    0.0178
    0.0156
    0.0138
    0.0123
    0.0111
    0.0100
    0.0091

inverse_square_radiust =

   1.0e+06 *

    1.0000
    0.4444
    0.2500
    0.1600
    0.1111
    0.0816
    0.0625
    0.0494
    0.0400
    0.0331
    0.0278
    0.0237
    0.0204
    0.0178
    0.0156
    0.0138
    0.0123
    0.0111
    0.0100
    0.0091

time =

   39.1100
   16.9700
    9.6100
    7.0400
    4.8100
    3.5100
    2.6600
    2.3000
    1.9100
    1.6000
    1.4800
    1.3200
    1.0200
    0.9400
    0.8400
    0.8200
    0.7700
    0.7200
    0.6600
    0.6300

timet =

   31.6000
   14.0400
    7.9000
    5.0600
    3.5100
    2.5800
    1.9800
    1.5600
    1.2600
    1.0400
    0.8800
    0.7500
    0.6400
    0.5600
    0.4900
    0.4400
    0.3900
    0.3500
    0.3200
    0.2900

er =

    0.6700
    0.4600
    0.2400
    0.2800
    0.2200
    0.3700
    0.3400
    0.2000
    0.1700
    0.1900
    0.2400
    0.2100
    0.1800
    0.1900
    0.2300
    0.1800
    0.1700
    0.1800
    0.1700
    0.1800

ert =

    0.9500
    0.4200
    0.2400
    0.1500
    0.1100
    0.0800
    0.0600
    0.0500
    0.0400
    0.0300
    0.0300
    0.0200
    0.0200
    0.0200
    0.0100
    0.0100
    0.0100
    0.0100
    0.0100
    0.0100

[Nicholas Pfaff]

This is the error message:

Operands to the || and && operators must be convertible to logical scalar values.

Error in legend (line 162)
        all(isgraphics(args{1})) % legend(children,strings,...)

Error in Linearized_plot_both (line 34)
legend([h,ht],'time vs. inverse_square_radius (Data)','Linearized Fit (Data)','time vs.
inverse_square_radius (Theory)','Linearized Fit (Theory)');

[dpb]

Please attach a mat or text file with the variables. Help us help you...

[Nicholas Pfaff]

I have attached it

Answer 1

[dpb]

h and ht are column vectors of handles and you've concatenated them into a 2x2 array instead of a vector...it's trying to put those together inside legend that's failing.

Use ; instead of , --

legend([h;ht],'time vs. inverse_square_radius (Data)','Linearized Fit (Data)','time vs. inverse_square_radius (Theory)','Linearized Fit (Theory)');

You'll probably cover up most of the plot area with the length of the text, but that's another problem... :)

[Nicholas Pfaff]

Thank you!

[dpb]

legend documents subset handles input must be a vector but seems rude for it to fail unceremoniously; it couldn't cost anything for it to use (:) on the argument inside to ensure is a vector or, failing that, inform of the what the error is during input check...

I didn't catch it just reading the code; set breakpoint in editor to see what you ended up with (hence the request for sample or data to run yours; almost certainly if had tried to emulate I'd have missed the crucial mistake and not found the problem).

ADDENDUM

I did submit enhancement SRQ to either add informative error or, it turns out that one can modify the failing line by the (:) and it will then accept the array of handles. I made the patch and recompiled the p-code and with that the existing code runs to completion with the expected result. No big deal but even an informative error would be better than the crash inside TMW-supplied function with no guidance as to why.