Unable to perform assignment because the left side is 1-by-3 and the right side is 1-by-2

Hello!

I am still struggling with trying to make graphs with categorical variables and error bars.

I am getting an error in line 11 (starts with ctr) that says "Unable to perform assignment because the left side is 1-by-3 and the right side is 1-by-2".

Am I missing a blatant 1-by-3 vector and how do I fix it without ruining my code? I apprieciate any help I can get, as I started learning Matlab just this year. Thanks!!

Here is my current code:

xc = categorical({'PTSD-','PTSD+'});
x = 1:numel(xc);
y = [.09 .13 ; -.08 -.35];
hBar = bar(x,y);
pos = [.1 .15 .2 .21];
neg = pos
posr = reshape(pos, 2, [2]);
negr = reshape(neg, 2, [2]);

for k1 = 1:size(y,2)
    ctr(k1,:) = bsxfun(@plus, 1:numel(hBar(k1).XData), hBar(k1).XOffset'); 
    ydt(k1,:) = hBar(k1).YData;
end

set(gca, 'XTick',x, 'XTickLabel',xc,'FontSize',12, 'FontName', 'Helvetica')

ylabel("Overall Cognitive Functioning (Z-Score)", 'FontSize',12, 'FontName', 'Helvetica')
ylim([-.5 .5]);
yticks([-.5, -.4 -.3 -.2 -.1 0 .1 .2 .3 .4 .5])

legend('location','northeast')
legend('FontName', 'Helvetica', 'FontSize',12)
legend boxoff

box off
legend boxoff

newcolors = ['#CBCBCB'; '#0033A0']
colororder(newcolors)

legend({'Met-','Met+'})
set(gca, 'XTick',x, 'XTickLabel',xc,'FontSize',12, 'FontName', 'Helvetica')

Answer 1

[Star Strider]

I’m not certain what you’re doing with that line.

The standard way of coding it (to put errorbar objects or text objects at the top of the bars) is:

ctr(k1,:) = bsxfun(@plus, hBar(k1).XData, hBar(k1).XOffset')

I suspect that diference could be the problem, however I find it a bit difficult to follow your code, because I don’t see that you use ‘ctr’ or ‘ydt’ anywhere else.

[Colton Rippey]

Unfortunately, I got the same error code. Honestly, I have a hard time following my code as well.

Thanks for your suggestion though!

[Star Strider]

You may have used the same block of code earlier, and set ‘ctr’ to a different size. The only fix for that I can think of is to re-define ‘ctr’ and ’ydt’ before the loop. The only other problem I can see is that the loop should be defined by ‘size(y,1)’ to make it appropriatlely general (although with a square matrix, that problem would not be evident).

Try this:

xc = categorical({'PTSD-','PTSD+'});
x = 1:numel(xc);
x = 1:2;
y = [.09 .13 ; -.08 -.35];
hBar = bar(x,y);
pos = [.1 .15 .2 .21];
neg = pos;
posr = reshape(pos, 2, [2]);
negr = reshape(neg, 2, [2]);

ctr = zeros(size(y));                                               % <— ADD THIS PREALLOCATION
ydt = zeros(size(y));                                               % <— ADD THIS PREALLOCATION

for k1 = 1:size(y,1)
    ctr(k1,:) = bsxfun(@plus, hBar(k1).XData, hBar(k1).XOffset');   % Note: ‘XOffset’ Is An Undocumented Feature, This Selects The ‘bar’ Centres
    ydt(k1,:) = hBar(k1).YData;                                     % Individual Bar Heights
end

If my diagnosis is correct, the preallocations should be the cure!

[Colton Rippey]

You are a damn legend, Star Strider. Thanks for your help!

[Star Strider]

As always, my pleasure!