pwGoodnessOfFit


 [LL, AIC, AICc, BIC, pValue1, pValue2] = pwGoodnessOfFit(chisq, N, p)

 Calculates several measures describing the goodness of fit.

 

Arguments for pwGoodnessOfFit

 chisq    Chi-Square value
 N        Number of fitted data points
 p        Number of fitted parameters

 LL       Log-Likelihood = - 0.5 * chisq - N/2 * log(2*pi)
 AIC      Akaike Information Criterion = - 2 * LL + 2 * p
 AICc     Corrected AIC = AIC + 2 * p * (p+1) / (N - p - 1)
 BIC      Bayesian Information Criterion = -2 * LL + p * log(N)
 pValue1  pValue(N)   = 1 - Cdf(chisq, N)
 pValue2  pValue(N-p) = 1 - Cdf(chis, N-p)

 

Description

 Cdf(.,f) denotes the cumulative density function of
 a chi-square distribution with f degrees of freedom.

See also

pwGetChisqAndN