My imagesc plot seems to be inverted

8 次查看（过去 30 天）

fadams18 2022-2-24

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1657535-my-imagesc-plot-seems-to-be-inverted

评论： Rik 2022-2-24

Below is my code. and the attached plot .

Mloops = [1,10:10:90];
Missing=10:10:90;
for e = 1:length(Mloops)
    eta =Mloops(e);     
    for i=1:length(Missing)
        mv= Missing(i);
        
        load( ['res/res_',int2str(eta),'_',int2str(mv),'.mat'] ,'RRE_NeA');
        
         sumRRE(i) = mean( RRE_NeA(RRE_NeA< 1e-5));
  
    end  
    RRE(:,e) = sumRRE;
    clear summRRE;  
end
eta = [1,10:10:90];
mv = [10:10:90];
imagesc(eta, mv,RRE); colorbar()
set(gca, 'YDir','normal')
xlabel('Maxloops'); ylabel(' MV ');

I expect to have a better error for Maxloops values > 30 ish. but it seems to me, its the inverse it is showing. my raw values shows that my results are correct but the plot doesnt.

5 个评论
显示 3更早的评论隐藏 3更早的评论

fadams18 2022-2-24

Its the relative reconstruction error. im doing some low rank approximations. So a high error means low value. so it should be around the blue. if the value is very small. meaning the error is low. so it should be around the yellow. and my experiment is such that, for high values of the eta( x-axis) and low values of MV (y-axis) I expect the error to become very low hence orange-yellow colors.