MATLAB Answers


Undefined function 'plus' for input arguments of type 'matlab.ui.Figure'.

Asked by Daisy
on 7 Oct 2014
Latest activity Edited by Matthew
on 5 May 2015
I get this error message when I run a script that was given to me on my data and may well have been written in an earlier version - I am working 2014b.
Any ideas? This is the full error message, but I haven't copied the script in as it's really quite long.
Undefined function 'plus' for input arguments of type 'matlab.ui.Figure'.
Error in PRL_analyseData>getNewFig (line 538) hfig = varargin{1}+1;
Error in PRL_analyseData (line 160) hfig = getNewFig(hfig,cmap);

  1 Comment

You probably need to rework the script to use an array of figure objects rather than doing maths on the old handles to refer to presumably the next figure.

Sign in to comment.

3 Answers

Answer by Guillaume
on 7 Oct 2014
 Accepted Answer

The reason it does not work in 2014b is because figures handles are no longer numbers. Thus you can't add 1 to them anymore.
Without seeing anymore of the code, it's difficult to give you any advice on how to solve this.


Follow up to the misplaced comment/answer:
Replace the first line of
for t = 1:hfig
for t = 1:hfig.Number
If you're happy with the solution provided, you should accept it.

Sign in to comment.

Answer by Daisy
on 8 Oct 2014

Many thanks, that has definitely helped. However I now have a different error message:
Undefined function 'vline' for input arguments of type 'double'.
Error in PRL_analyseData (line 266)
This is the section of the code it relates to. Is it to do with 2014b changes again?
% average and smoothed average probability of chosing correctly on each
% trial
choiceBinary = choice;
plotdata{1} = choiceBinary;
plotdata{2}= mySmooth(choiceBinary,span,2);
for n = 2 % can also plot the raw data, then n=1
hfig = getNewFig(hfig,cmap);
mu = []; sem = [];
for t = cond
mu(:,t) = mean(plotdata{n}(groupMember==t,:));
switch n
case 1
case 2
sem(:,t) = std(plotdata{n}(groupMember==t,:),[],1)/sqrt(nsgroup(t)); end
if n==2
b = nan(nt,1,max(cond));
b(:,1,:) = sem;
x = repmat(1:nt,max(cond),1)';
ylim([0 1])


Show 1 older comment
This I've actually managed to resolve with further investigation. Thanks
My apologies. Yes, I worked out that vline came from the file exchange. Thanks. I now have an issue with:
No method 'colon' with matching signature found.
Error in PRL_analyseData (line 303)
for t = 1:hfig
This relates to this section of the script:
hfig = getNewFig(hfig,cmap);
mu = []; sem = [];
for t = cond
mu(1,t) = mean(WS(groupMember==t));
mu(2,t) = mean(LS(groupMember==t));
sem(1,t) = mean(WS(groupMember==t))/sqrt(nsgroup(t));
sem(2,t) = mean(LS(groupMember==t))/sqrt(nsgroup(t));
for t = 1:hfig
I am reading about signatures in order to try and resolve this, but would appreciate any help
Please, find the reply as a comment to my answer rather than here.

Sign in to comment.

Answer by Matthew
on 5 May 2015
Edited by Matthew
on 5 May 2015

An updated answer to this is to overload the 'plus' function. It seems that matlab calls the builtin for inline operators by default, even if the overload doesn't pay any attention to the rules for overloading a function. I have no idea why this is though, as overloading other critical functions like sprintf seems to cause a world of pain.
Regardless of the reason, this oddity means the following function can be placed anywhere in your path, and it will only be called if the regular plus function can't handle the inputs, so it doesn't impact normal behavior or performance.
function out = plus(A,B)
out = double(A) + double(B)
%Fail with the built-in


Sign in to comment.