How to avoid decreasing in length of file

I have been doing some manipulation in the for loop. I have constantly found that the calculation only works for some of the length of variables I assigned i.e. if jj = length(tcal) where tcal is 26, in the out put of results I usually find jj=1 or 2. what will be the possible reason and how can I avoid it . I can give you the whole script if needed.

Thanks for your help. It giving me hard time...

[Matt J]

If tcal is 26 (a scalar) then it makes a lot of sense to get jj=1,

>> jj=length(26)

jj =

       1

[Ede gerlderlands]

Here is the complete code

for jj = 1:length(tcal);

    calpre= HADCP.pressure(hmin:hmax);    
    x= cast(calpre, 'double');%%%changing data type
    kk =1:1:length(x);
    B=repmat(reshape(x(kk),1,1,length(x)),20,40);
    h2(:,:,jj)=h(:,:,jj)-B(:,:,jj).*0.001;      
    sig1 = repmat(sig,[1 1 26]);
    sig2(:,:,jj)=sig1(:,:,jj).*(h2(:,:,jj)./h(:,:,jj));
    C(:,:,jj) = max(sig2(:,:,jj),[],1);       
    Sca1l = permute(Sca1 ,[3 2 1]);    
                   Sca13(:,:,jj)=griddata(sig1(:,:,jj),bet,Sca1l(:,:,jj),C(1,:,jj),bet(1,:));   Sca13=squeeze(Sca13(1,1:17,jj)); 

end

[Matt J]

It doesn't run. There appears to be a fragmentary line

h3=h(end+ 2.*tan(0.06).*R

Also, it makes no sense if indeed tcal=26. As I told you length(tcal) is then equal 1 and jj will not loop.

[Ede gerlderlands]

h3=h(end+ 2.*tan(0.06).*R is commented line sorry I had to delete it. tcal has 26 data and not just 26. And there are some files which I didn't present. Yes , jj becomes 25 ,1 and 26 when I run in different modes and my question is why it changes this way and what can I do about it? thanks..