How to correspond line colors with histogram

Question

MadjeKoe 2020-10-15

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/614833-how-to-correspond-line-colors-with-histogram

评论： Star Strider 2020-10-15

Schermafbeelding 2020-10-15 om 13.42.45.png

I made these overlapping histograms using a for loop. I'm trying to get the lines in the same color as the histograms. Does anybody know how to do this?

[~,ErrorLength]= size(out);

for i =1:ErrorLength

hold on

m = histogram(out(:,i),'normalization','pdf');

f = fitdist(out(:,i),'kernel');

hoi = pdf(f,x);

l = line(x,hoi,'linestyle','-','linewidth',2);

hold off

end

xlabel('Orientation');

ylabel('Error rate');

legend('Target 1','','Target 2','','Target 3','','Target 4','')

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Star Strider 2020-10-15

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/614833-how-to-correspond-line-colors-with-histogram#answer_514713

编辑：Star Strider 2020-10-15

在 MATLAB Online 中打开

The FaceColor property of the bars is 'auto', so if you want the colors to be the same, it is necessary to define them:

out = randn(500,3).*[10 20 30];                                         % Create Data To Test Code
x = linspace(-80, 80);                                                  % Create Data To Test Code
cm = jet(size(out,2));                                                  % Define ‘colormap’
[~,ErrorLength]= size(out);
for i =1:ErrorLength
    hold on
    m = histogram(out(:,i),'normalization','pdf')
    m.FaceColor = cm(i,:);                                              % Define ‘FaceColor’ 
    f = fitdist(out(:,i),'kernel');
    hoi = pdf(f,x);
    l = line(x,hoi,'linestyle','-','linewidth',2, 'Color',cm(i,:));     % Define Line ‘Color’ 
    hold off
end
xlabel('Orientation');
ylabel('Error rate');
legend('Target 1','','Target 2','','Target 3','','Target 4','')

You will need to fix the problem with the legend.

EDIT — (15 Oct 2020 at 12:17)

Added ‘out’ and ‘x’ definitions.

8 个评论
显示 6更早的评论隐藏 6更早的评论

Star Strider 2020-10-15

在 MATLAB Online 中打开

No worries!

In that event, it is simply necessary to keep track of the histogram handles, then pass them to the legend call, and lightly edit the legend call arguments.

The full code is now:

out = randn(1500,4).*[10 20 30 40];                                     % Create Data To Test Code
x = linspace(-80, 80);                                                  % Create Data To Test Code
cm = jet(size(out,2));                                                  % Define ‘colormap’
[~,ErrorLength]= size(out);
figure
hold on
for i =1:ErrorLength
    m{i} = histogram(out(:,i),'normalization','pdf')
    m{i}.FaceColor = cm(i,:);                                           % Define ‘FaceColor’ 
    f = fitdist(out(:,i),'kernel');
    hoi = pdf(f,x);
    l = line(x,hoi,'linestyle','-','linewidth',2, 'Color',cm(i,:));     % Define Line ‘Color’ 
end
hold off
xlabel('Orientation');
ylabel('Error rate');
legend([m{:}], 'Target 1','Target 2','Target 3','Target 4')

Note that ‘m’ is now ‘m{i}’, and they are included as the first argument of the lightly edited legend call. (I also updated my version of your code to make it a bit more efficient.) The edge lines around the histogram bars have returned as well.

MadjeKoe 2020-10-15