Re: Problem with Lsode..help

[genehacker]

Hi Carlo,

  thanks very much. it works fine when i initialize x4 with a value
upfront:)


Carlo de Falco-2 wrote:
> 
> 
> On 27/ott/08, at 21:28, genehacker wrote:
> 
>>
>> Hi Bryan, Here is the script
>>
>> **********************
>>
>> clf;
>> clc;
>> clear all;
> 
>> 1;
> This command makes no sense, why is it there??
> 
> right. i need to remove this. I put it before so as to treat the file not
> as a function. now it doesnt make sense
> 
>> global x4;
> this does not assign a value to x4
> 
>> x5 = @(x1)   k1max- ((k1max- k1min)/(1 + (f1 * x1/IC2)));
>> x6 = @(x1)   k2max- ((k2max- k2min)/(1 + (0.01*f1 * x1/IC2)));
>> x7 = @(x1)   k23min- ((k23min- k23max)/(1 + (0.01*f1 * x1/IC2)));
>> t0 = 8;
>>
>> x8 = @(t,x2) ((t+0.001) .* (t<t0)) + (t0 .*(t==t0)) +
>> (t0*exp(0.1*t0)*exp(-0.1*t) .*(t > t0));
>> xdot = @(x,t) [ ( (x8(t,x(2)) * x4/pre_OC) - x5(x(1)) * x(1));
>>                      (x6(x(1)) * pre_OB - x7(x(1)) * x(2));
>>                      (x7(x(1)) * x(2) - k3 * x(3)) ];
> the definition of xdot contains x4, so x4 should be initialized before
> xdot is used
>>
>>
>> x = [0,0,0; 0,0,0];
>> len = 100;
>> time = linspace(0,len,len+1);
>>
>> matrix = zeros(1,10);
>> matrix(1,10) = bmd = 100; % for bone initial state
>> a = 0.06;    % OB weight
>> b = 0.1;     % OC weight
>> for i=1:(length(time)-1)
>>      seedtime =  time(i:i+1);
>>      [x,istate,msg] = lsode(xdot,[x(2,1),x(2,2),x(2,3)],seedtime);
> lsode will repeatedly call the function xdot but, as x4 is not yet  
> defined, xdot(1)
> will evaluate to the empty matrix [] so xdot(x,t) will be a 2-vector.
> 
> 
>>
>>       bmd = bmd + a*x(2,3) - b*x(2,1);
>>      x4 = (vmax * pre_OC) / ( x8(time(i+1),x(2,2)) * (pre_OC + (x8(time(i 
>> +1),x(2,2)) * (1 + (10 * x(2,2) / kI))  )) );
> you won't get to this point as your script crashes on calling lsode  
> above, but this is also wrong. lsode as you called it above will  
> return a 3-vector which will overwrite the 2x3 matrix x
> so x(2,2) will not exist after lsode returns. Plus vmax, pre_OC are  
> undefined.
> x would be a 2x3 vector since i made seedtime a 2-vector. vmax, pre_oc are
> defined at the beginning
>>
>>      matrix(i+1,:) = [time(i+1) x8(time(i+1),x(2,2)) x(2,1) x(2,2) x(2,3)
>> x5(x(2,1)) x6(x(2,1)) x7(x(2,1)) x4 bmd];
>> endfor
>>
>> timerescale = 1;
>> plot(time/timerescale,10*matrix(:,3),"r"); hold on ;
>> plot(time/timerescale,10*matrix(:,5),"m"); hold on;
>> plot(time/timerescale,matrix(:,10),"b"); hold off;
>> xlabel("DAYS");
>> ylabel('Concentration');
>> axis([0 len]);
>> legend('OC','OB','bmd');
>> print -djpg foo.jpg;
>>
>> *************************
>>
> 
> c.
> `
> _______________________________________________
> Help-octave mailing list
> address@hidden
> https://www-old.cae.wisc.edu/mailman/listinfo/help-octave
> 
> 

-- 
View this message in context: 
http://www.nabble.com/Problem-with-Lsode..help-tp20193330p20198429.html
Sent from the Octave - General mailing list archive at Nabble.com.