next up previous
Next: 2.5 Residuals <filename> Up: 2 Commands Previous: 2.3 <Parameter> = <value>

2.4 Fit

You can now fit the model to the data. This will probably take quite a bit of time, especially if there is a large number of parameters to estimate or if the data set is large. The computer time increases with the sample size to the power of 2.

The program calculates the maximum likelihood estimates of each parameter, and the maximum likelihood $\ln L$. In addition the expectations, variances, correlation of X and Y, EX, EY, $\operatorname{Var}(X)$, $\operatorname{Var}(Y)$ and $\rho(X,Y)$, of the fitted dispersal distribution are calculated. These dispersion measures enters evolutionary models of spread of advantegeous genes, models of genetic variation in migration-selection clines, etc. In addition to this, the number of residual degrees of freedom and the estimated amount of overdispersion $\widehat{c}$ are calculated.