Res ni narejena za numeriko. matlab bi se tukaj bolje odrezal, se najbolje pa je uspesen gnuplot, ki nafita z eno kratko ukazno vrstico in ti vrne vse te parametre zastonj.
Če je pa treba kej več (histogrami pridejo takoj na misel) pa pač nardim v gnuplotu/mathematici.Koda: Izberi vse
1 2 10
2 4 10
3 6 10
4 8 10
5 10 10
6 12 10
7 14 10
8 16 10Koda: Izberi vse
gnuplot> fit z=a*x+b 'test.dat' using 1:2:3 via a,b
Iteration 0
WSSR : 1.68879e-028 delta(WSSR)/WSSR : 0
delta(WSSR) : 0 limit for stopping : 1e-005
lambda : 0.357071
initial set of free parameter values
a = 2
b = 1.01337e-013
*********************
After 1 iterations the fit converged.
final sum of squares of residuals : 1.68879e-028
rel. change during last iteration : 0
degrees of freedom (FIT_NDF) : 6
rms of residuals (FIT_STDFIT) = sqrt(WSSR/ndf) : 5.30533e-015
variance of residuals (reduced chisquare) = WSSR/ndf : 2.81466e-029
Singular matrix in Invert_RtR
error during fitTorej, izognejo se direktnemu porocanju napake, ker to velja samo, ce je fit tocen (ce so napake porazdeljene nakljucno). Ce prav razumem, so porocane napake (asymptotic standard error) tocne, ce je hi-kvadrat blizu 1. Ce ni, potem je temu primerno vse skalirano.At the conclusion fit reports ’stdfit’, the standard deviation of the fit, which is the rms of the residuals, and the variance of the residuals, also called ’reduced chisquare’ when the data points are weighted. The number of degrees of freedom (the number of data points minus the number of fitted parameters) is used in these estimates because the parameters used in calculating the residuals of the datapoints were obtained from the same data. These values are exported to the variables
FIT_NDF = Number of degrees of freedom
FIT_WSSR = Weighted sum-of-squares residual
FIT_STDFIT = sqrt(WSSR/NDF)
To estimate confidence levels for the parameters, one can use the minimum chisquare obtained from the fit and chisquare statistics to determine the value of chisquare corresponding to the desired confidence level, but considerably more calculation is required to determine the combinations of parameters which produce such values.
Rather than determine confidence intervals, fit reports parameter error estimates which are readily obtained from the variance-covariance matrix after the final iteration. By convention, these estimates are called "standard errors" or "asymptotic standard errors", since they are calculated in the same way as the standard errors (standard deviation of each parameter) of a linear least-squares problem, even though the statistical conditions for designating the quantity calculated to be a standard deviation are not generally valid for the NLLS problem. The asymptotic standard errors are generally over-optimistic and should not be used for determining confidence levels, but are useful for qualitative purposes.