Below I have pasted the outputs of an ordistep() on an rda()
I noticed in the final fitting of the ordistep() that two terms
"mydata$TreeCov" and "mydata$HabConfig" were still included but had variance of
0 and therefore P was not calculated.
Why are they still included in the final model?
thanks
Nevil Amos
> myOrdistepBoth
Call: rda(formula = mygenind@tab ~ mds3dCS_NULL + mds3dTRE_25_2_CS25 +
mydata$LAT.x + mydata$Hab_Config + mds3dTRE_25_100_CS25 + mydata$TreeCov +
mydata$Site_No + mds3dSFW_EO_100_CS25 + mds3dSFW_EO_5000_CS25 +
mydata$Landscape + mds3dSFW_TH_10_CS25 + mydata$LONG.x, scale = T, na.action =
"na.omit")
Inertia Proportion Rank
Total 160.0000 1.0000
Constrained 45.3560 0.2835 32
Unconstrained 114.6440 0.7165 139
Inertia is correlations
Some constraints were aliased because they were collinear (redundant)
Eigenvalues for constrained axes:
RDA1 RDA2 RDA3 RDA4 RDA5 RDA6 RDA7 RDA8 RDA9 RDA10 RDA11
RDA12 RDA13 RDA14 RDA15 RDA16 RDA17 RDA18 RDA19 RDA20 RDA21 RDA22
RDA23 RDA24 RDA25
3.3489 3.2291 2.8211 2.5129 2.2606 2.1062 2.0834 1.8720 1.7727 1.6680 1.6072
1.5512 1.5100 1.4314 1.3802 1.2427 1.1497 1.1153 1.0547 1.0287 0.9814 0.9361
0.9168 0.8851 0.7577
RDA26 RDA27 RDA28 RDA29 RDA30 RDA31 RDA32
0.7150 0.7056 0.6436 0.6098 0.5643 0.4806 0.4140
Eigenvalues for unconstrained axes:
PC1 PC2 PC3 PC4 PC5 PC6 PC7 PC8
3.433 3.098 3.057 2.912 2.866 2.743 2.669 2.586
(Showed only 8 of all 139 unconstrained eigenvalues)
> anova(myOrdistepBoth,by="term")
Permutation test for rda under reduced model
Terms added sequentially (first to last)
Model: rda(formula = mygenind@tab ~ mds3dCS_NULL + mds3dTRE_25_2_CS25 +
mydata$LAT.x + mydata$Hab_Config + mds3dTRE_25_100_CS25 + mydata$TreeCov +
mydata$Site_No + mds3dSFW_EO_100_CS25 + mds3dSFW_EO_5000_CS25 +
mydata$Landscape + mds3dSFW_TH_10_CS25 + mydata$LONG.x, scale = T, na.action =
"na.omit")
Df Var F N.Perm Pr(>F)
mds3dCS_NULL 3 6.527 2.6380 99 0.01 **
mds3dTRE_25_2_CS25 3 5.596 2.2616 99 0.01 **
mydata$LAT.x 1 1.892 2.2939 99 0.01 **
mydata$Hab_Config 2 2.858 1.7325 99 0.01 **
mds3dTRE_25_100_CS25 3 4.314 1.7435 99 0.01 **
mydata$TreeCov 1 1.238 1.5008 99 0.01 **
mydata$Site_No 1 1.304 1.5806 99 0.01 **
mds3dSFW_EO_100_CS25 3 3.834 1.5497 99 0.01 **
mds3dSFW_EO_5000_CS25 3 3.344 1.3515 99 0.01 **
mydata$Landscape 8 9.711 1.4717 99 0.01 **
mds3dSFW_TH_10_CS25 3 3.702 1.4961 99 0.01 **
mydata$LONG.x 1 1.037 1.2569 99 0.06 .
Residual 139 114.644
---
Signif. codes: 0 *** 0.001 ** 0.01 * 0.05 . 0.1 1
> anova(myOrdistepBoth,by="margin")
Permutation test for rda under reduced model
Marginal effects of terms
Model: rda(formula = mygenind@tab ~ mds3dCS_NULL + mds3dTRE_25_2_CS25 +
mydata$LAT.x + mydata$Hab_Config + mds3dTRE_25_100_CS25 + mydata$TreeCov +
mydata$Site_No + mds3dSFW_EO_100_CS25 + mds3dSFW_EO_5000_CS25 +
mydata$Landscape + mds3dSFW_TH_10_CS25 + mydata$LONG.x, scale = T, na.action =
"na.omit")
Df Var F N.Perm Pr(>F)
mds3dCS_NULL 3 3.519 1.4222 199 0.00500 **
mds3dTRE_25_2_CS25 3 3.663 1.4805 199 0.00500 **
mydata$LAT.x 1 1.185 1.4364 199 0.00500 **
mydata$Hab_Config 0 0.000 0.0000 0
mds3dTRE_25_100_CS25 3 3.569 1.4423 199 0.00500 **
mydata$TreeCov 0 0.000 0.0000 0
mydata$Site_No 1 1.251 1.5170 199 0.00500 **
mds3dSFW_EO_100_CS25 3 3.303 1.3348 199 0.00500 **
mds3dSFW_EO_5000_CS25 3 3.395 1.3721 199 0.01000 **
mydata$Landscape 8 9.573 1.4508 199 0.00500 **
mds3dSFW_TH_10_CS25 3 3.658 1.4785 199 0.00500 **
mydata$LONG.x 1 1.037 1.2569 1399 0.03571 *
Residual 139 114.644
---
Signif. codes: 0 *** 0.001 ** 0.01 * 0.05 . 0.1 1
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
