Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
NCTM or The Math Forum.



Re: Non Linear Regression.
Posted:
Sep 25, 2013 3:27 PM


On 9/25/2013 1:31 PM, dpb wrote: > On 9/25/2013 10:03 AM, satyaraj wrote: >> Hi everyone. Sorry i am very new to matlab and to nonlinear regression. >> >> This is what i would like to do, i have the formula as stated below. >> >> E=Summation of n (i=1) * [ yi F(8,xi)] ^2 >> >> E = objective function F = reservoir model function 8 = unknown >> reservoir parameters 2; = dependent variable (time) y; = independent >> variable (pressure) n = number of data >> The parameters 8, could be from 1 to 5 of them. Which each parameter has >> their own formula linking one another. >> such as tD = (0.000264kt) / >> (porosity*viscosity*compressibility*radius^2) >> *note : only k and t is not available. other values are. >> So, when i supply huge amount of data for the time and pressure, t p, i >> would like to calculate the parameters via iterations. would that be >> possible? > > Sure....how best depends significantly on which toolboxes you have. > > If you have the Statistics Toolbox > > doc nlinfit > doc nlintool > doc nonlinearmodel > > If not, you can use the other minimization tools and write the > objective function as above and solve with a little more detail needed > to be included explicitly. > > 
In addition, if you have Optimization Toolbox there are the lsqnonlin and lsqcurvefit functions.
Alan Weiss MATLAB mathematical toolbox documentation



