removing NaN values from my codes

Hi, I have these codes:

result=zeros(1,12) avedata=mean(turkeyreturn); vardata=var(turkeyreturn); N=length(turkeyreturn); for h=1:14;

A=TRBB(:,h);

BB=[A turkeyreturn A.*turkeyreturn ];

% FF=BB(0== sum(isnan(BB), 2), :);

%R(:,h)=BB((100:end),2);

y1=0;y2=0; z1=length(find(A==1)); z2=length(find(A==-1));

for k=1:length(BB) if BB(k,1)==1 y1=BB(k,3)+y1; elseif BB(k,1)==-1 y2=BB(k,2)+y2; end end

for n=length(BB)

% BB=BB(0== sum(isnan(BB), 2), :); BB=BB(isfinite(BB(:, 1)), :) column=(BB(1:end,2)) summ= sum(BB(1:end,2)) av=summ/n annualav=av*256 standard=std(column) sharpe=((annualav-rf)/standard) end

aveofones=y1/z1; aveofmione=y2/z2; buy_sell=(aveofones-aveofmione);

T_one=(aveofones-avedata)/(sqrt((vardata/N)+(vardata/z1))); T_min1=(aveofmione-avedata)/(sqrt((vardata/N)+(vardata/z2))); T_both=(aveofones-aveofmione)/(sqrt((vardata/z1)+(vardata/z1)));

% R(:,h)=FF(:,3); result(h,:)=[z1,z2,aveofones,aveofmione,buy_sell,T_one,T_min1,T_both,av,annualav, standard,sharpe]

end

how do I save the the second col um BB matrix so that I can get all the results in for loap. additionally there are some NaN values in the third column. how can I remove these row with matching cell in the first and second column please from BB matrix and save them after all.The number of NaN is different in each loap.

thanks jean

[Image Analyst]

Don't double space, it's must simpler than that: just highlight and click {}Code. http://www.mathworks.com/matlabcentral/answers/13205-tutorial-how-to-format-your-question-with-markup

[jean]

what yo[u meant by double space please and where i did it?

[Andreas Goser]

It means your code is not formatted and thus difficult to help you.

[jean]

okay thanks will do it again now

[jean]

     Hi, I have these codes:

      result=zeros(1,12) 
    acedata=mean(turkeyreturn); 
    vardata=var(turkeyreturn); 
    N=length(turkeyreturn); 
    for h=1:14;

      A=TRBB(:,h);

      BB=[A turkeyreturn A.*turkeyreturn ];

      % FF=BB(0== sum(isnan(BB), 2), :);

      %R(:,h)=BB((100:end),2);

      y1=0;y2=0; 
      z1=length(find(A==1)); 
      z2=length(find(A==-1));

      for k=1:length(BB) 
      if BB(k,1)==1 y1=BB(k,3)+y1; 
      elseif BB(k,1)==-1 
      y2=BB(k,2)+y2; 
      end 
      end

      for n=length(BB)

      % BB=BB(0== sum(isnan(BB), 2), :); 
      BB=BB(isfinite(BB(:, 1)), :) 
      column=(BB(1:end,2)) 
      summ= sum(BB(1:end,2)) 
      v=summ/n 
      annualav=av*256 
      standard=std(column) 
      sharpe=((annualav-rf)/standard) 
      end

      aveofones=y1/z1; 
      aveofmione=y2/z2; 
      buy_sell=(aveofones-aveofmione);

      T_one=(aveofones-avedata)/(sqrt((vardata/N)+(vardata/z1))); 
      T_min1=(aveofmione-avedata)/(sqrt((vardata/N)+(vardata/z2)));
       T_both=(aveofones-aveofmione)/(sqrt((vardata/z1)+(vardata/z1)));

      % R(:,h)=FF(:,3); result(h,:)=[z1,z2,aveofones,aveofmione,buy_sell,T_one,T_min1,T_both,av,annualav, standard,sharpe]

      end

      how do I save the the second col um BB matrix so that I can get all the results in for loap. additionally there are some NaN values in the third column. how can I remove these row with matching cell in the first and second column please from BB matrix and save them after all.The number of NaN is different in each loap.

      thanks jean

Answer 1

[Andreas Goser]

Without being able to understand your code and assuming you want to remove nans from the data and not your code, here is an example:

a=[1 2 nan 4 5]
x=isnan(a)
a=a(~x)

[per isakson]

.. or

    a( isnan(a) ) = [];

[jean]

should i put it at the end of my codes. Need to remove the Nan and the value that match them in other two column.

[jean]

and want to save all the vales after that in one matrix and it saying matrix dimision does not match. I think the number of NaN is different from one column to another.