How to resolve this error ?

Question

Rashmi 2018-8-14

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/414800-how-to-resolve-this-error

回答： Adam Danz 2018-8-14

 z=imread('demo.jpg');
   r=z(:,:,1);
   g=z(:,:,2);
   b=z(:,:,3);
CREATE A TRAINING SET
% Identify the location (row,column) of "training pixels" and assign them a "group number":
% You will need to identify at least 2 training pixels per group
%
   tpix=[1309,640 ,1;... % Group 1: Yellow Bat
         1218,755 ,1;... 
         1351,1409,2;... % Group 2: Grey Concrete
         673 ,394 ,2;...
         285 ,1762,3;... % Group 3: Red Tub
   177 ,1542,3;...
   538 ,1754,4;... % Group 4: Blue Tub
   432 ,1811,4;...
   1417,2010,5;... % Group 5: White Tub
   163 ,1733,5;...
   652 ,677 ,6;... % Group 6: Baby
   864 ,1032,6;...
   836 ,1882,7;... % Group 7: Dark Doormat
   1063,1917,7]; 
   row=tpix(:,1);   % y-value
   col=tpix(:,2);   % x-value
   group=tpix(:,3); % group number
   ngroup=(group);
%
% From these pixels, make a training set consisting of each training pixel's band values
% "train" should have the same number of rows as the number of training pixels, and the
% same number of columns as number of bands (in this case 3, in Lansat case 7).  
%   
   train=[];
%    train=zeros(ngroup);
  % train=ones(group); %(1,ngroup);

tic [class,err,misfit]=classify(AllPix,train,row); toc

When I run this code , error has occured as "TRAINING must have more observations than the number of groups"

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

Adam Danz 2018-8-14

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/414800-how-to-resolve-this-error#answer_332718

在 MATLAB Online 中打开

Read and understand how classify() works

doc classify

The 2nd input to classify() is 'training' which is a grouping variable and as the error tells you, you must have more observations (number of elements in 'training' vector) than groups (number of unique values in 'training').

Your code defines 'training' as

train = [];

which is empty.