[R] Degrees of Freedom GAM/GAMM

Worthington, Thomas A thomas.worthington at okstate.edu
Mon May 5 22:29:30 CEST 2014


Dear All 

I've fitted a GAMM to relate water temperature to the day of the year (DOY) at three different sites. I used summary(MFinal$gam) and anova(MFinal$gam) to produce the output of my model. I'm confused on how to report the degrees of freedom for the smother and the factor. I currently have 

"The final model suggested water temperature was significant related to DOY (6.46 estimated degrees of freedom, F = 224.2, p < 0.001) and was significantly different between sites (F2 = 82.14, p <0.001)"

But think there should be more to the degrees of freedom particularly for the sites factor.

If anyone could advise on the correct way to display the results I would greatly appreciate it . The output from the model is below

Best wishes

Tom 


> summary(MFinal$gam)

Family: gaussian 
Link function: identity 

Formula:
Temperature ~ s(DOY) + Site

Parametric coefficients:
            Estimate Std. Error t value Pr(>|t|)    
(Intercept)  12.0366     0.2411   49.92   <2e-16 ***
SiteG         4.2460     0.3979   10.67   <2e-16 ***
SiteH         4.1439     0.4069   10.19   <2e-16 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

Approximate significance of smooth terms:
         edf Ref.df     F p-value    
s(DOY) 6.457  6.457 224.2  <2e-16 ***
---
Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1

R-sq.(adj) =  0.914  Scale est. = 0.098602  n = 137


> anova(MFinal$gam)

Family: gaussian 
Link function: identity 

Formula:
Temperature ~ s(DOY) + Site

Parametric Terms:
     df     F p-value
Site  2 82.14  <2e-16

Approximate significance of smooth terms:
         edf Ref.df     F p-value
s(DOY) 6.457  6.457 224.2  <2e-16 



More information about the R-help mailing list