how to plot data with small values

I want to plot some result which shows how different is the model A from model B etc. My error is something zero. do you have any idea how I can show that the interpretation would be easy too?

  Err 1  Err2  Err3     Err4  Err5
  A  0.0142   0       0.0410    0     0
  B  0  0  0.0138    0     0
  C  0  0  0         0     0
  D  0  0  0         0     0
  E  0.0142  0      0    0     0
  F  0  0  0         0     0
  G  0  0  0         0     0
  H   0.0135 0   0.0416667 0  0.0138889

Answer 1

[Star Strider]

I have no idea what you are doing, but the bar3 function might be what you want.

For example:

Xt =   [{'Err1'}  {'Err2'}  {'Err3'}   {'Err4'}  {'Err5'}];
Yt = [{'A'} {'B'} {'C'} {'D'}  {'E'}  {'F'}  {'G'}  {'H'}];
  M = [  0.0142   0       0.0410    0     0
    0  0  0.0138    0     0
    0  0  0         0     0
    0  0  0         0     0
    0.0142  0      0    0     0
    0  0  0         0     0
    0  0  0         0     0
     0.0135 0   0.0416667 0  0.0138889]; 

figure(1)
bar3(M)
set(gca, 'XTickLabel', Xt)
set(gca, 'YTickLabel', Yt)

produces:

[Niki]

this is a classification problem. The error for the first class. for the second class etc. I want to link them together using method A then it can be comparable with method B etc

[Star Strider]

I have no idea what you want. This is the only way I can think of to plot the data you presented.

[Niki]

Thanks! lets see if anyone else give me an idea, if not, Iwill accept your suggestion

[Niki]

I did try to plot it as you said

Xt=[{'CCRC1'} {'CCRC2'} {'CCRC3'} {'CCRC4'} {'CPRC1'} {'CPRC2'} {'CPRC3'} {'CPRC4'}]; Yt=[{'A'} {'B'} {'C'} {'D'} {'E'} {'F'} {'G'} {'H'} {'I'} {'AF'} {'AFI'} {'AG'} {'AGI'} {'BF'} {'BFI'} {'BG'} {'BGI'} {'CF'} {'CFI'} {'CG'} {'CGI'} {'DF'} {'DFI'} {'DG'} {'DGI'} {'EF'} {'EFI'} {'EG'} {'EGI'} {'HF'} {'HFI'} {'HG'} {'HGI'} {'MNCN'} {'FI'} {'GI'}]; figure (1) bar3(M) set(gca,'XtickLabel', Xt) set (gca, 'YtickLabel', Yt)

But seems not working very well. if you look at the X axis , it does not show everything . On the other hand, the y axis is very narrow and cause problem. Do you have any suggestion? You can find the results in attachment

[Star Strider]

I have problems opening your ‘.xls’ file, since it is mixed string and numeric. I simulated your matrix with a same-size random matrix.

Try this:

set(gca, 'XtickLabel', Xt, 'FontSize',6) 
set (gca, 'YTick', 1:36, 'YtickLabel', Yt, 'FontSize',6)
view([-45  45])

[Niki]

Did not work,

Here is the data in the attachment

[Star Strider]

What doesn’t work? It looks good to me.

[Niki]

You are right.I made mistake I used the previous "set" and then over right the last one. anyway I found the same figure as you did. I have one question before I accept your solution. can you please let me know whether you know how to plot it like following figure? because if you see your plot does not show for the last one (one columns before the last is dominating the other) . do you know how to plot it like attached file ?

[Star Strider]

The ribbon plot is about as close as we can get. You can change the width of the ribbons and rotate the plot to get the result you want:

[X,Y] = meshgrid(1:8, 1:36);

figure(1)
ribbon(Y, M, 0.1)
set(gca, 'Xtick', 1:8, 'XtickLabel', Xt, 'FontSize',6) 
set (gca, 'YTick', 1:36, 'YtickLabel', Yt, 'FontSize',6)
view([50 50])

The Xt and Yt cell arrays stay the same. The meshgrid call is new, and the 0.1 is the ribbon width. Otherwise the data are the same.