MATLAB: Matlab doesn’t fit :(

fitwoods saxon

I'm trying to generate a fittype to fit my data, but matlab only plot a y=0 function. I don't know what to do, i've tried all. This is my code
x=[369  370  371  372  373  374  375  376  377  378  379  380  381  382  383  384  385  386  387  388  389  390  391  392  393  394  395  396  397  398  399  400  401  402  403  404  405  406  407  408  409  410  411  412  413  414  415  416  417  418  419  420  421  422  423  424  425  426  427  428  429  430  431  432  433  434  435  436  437  438  439  440  441  442  443  444  445  446  447  448  449];
y=[26  26,7500000000000  31,5000000000000  28,2500000000000  28,7500000000000  39,5000000000000  30  32  25,2500000000000  36,2500000000000  22,5000000000000  28  39,2500000000000  41,2500000000000  38,7500000000000  30,7500000000000  24,5000000000000  32,7500000000000  23,5000000000000  25,5000000000000  24,7500000000000  30  31  23,2500000000000  25  30,2500000000000  19,2500000000000  24  23,7500000000000  24,2500000000000  22,2500000000000  25,7500000000000  13,2500000000000  11  23  10  14,5000000000000  21  17  15,2500000000000  19,7500000000000  17  18  14,7500000000000  12  11,2500000000000  14  5,75000000000000  7  13  8,25000000000000  6,75000000000000  11,5000000000000  9,25000000000000  3,50000000000000  11,5000000000000  3,50000000000000  8,25000000000000  6,50000000000000  -1  5  10,2500000000000  11,7500000000000  11,2500000000000  8  4  5  2,75000000000000  4,75000000000000  4,75000000000000  12,2500000000000  11  3,25000000000000  3,75000000000000  4,50000000000000  4,50000000000000  5,50000000000000  7,25000000000000  16  7,75000000000000  8,25000000000000];
myfittype=fittype('v/(1+exp((x-r)/a))','independent',{'x'},'dependent',{'y'},'coefficients',{'v','r','a'});
myfit=fit(x',y',myfittype,'DiffMinChange',0.0000001)
plot(myfit,'--r',x,y,'.k');
The function is a woods-saxon typus
I would be very happy if somebody can help me, I don't know what to do!:(

Best Answer

The random start point used is a terribly poor choice here. That is a common problem with ANY exponential variation of fit, as a random start point makes no sense, often resulting in overflows or underflows.
Supply an intelligent set of starting values. For example:
myfit=fit(x',y',myfittype,'DiffMinChange',0.0000001,'start',[25 360 20])
Related Question