adonis2 from version

[PierreQuevreux]

Hello,

I went back to a former analysis I did with a previous version of the package and the output of adonis2 is not the same. Here the example with the data set included in the package.

With version 2.6.4

> data(dune)
> data(dune.env)
> adonis2(dune ~ Management*A1, data=dune.env, permutations=99)
Permutation test for adonis under reduced model
Terms added sequentially (first to last)
Permutation: free
Number of permutations: 99

adonis2(formula = dune ~ Management * A1, data = dune.env, permutations = 99)
              Df SumOfSqs      R2      F Pr(>F)  
Management     3   1.4686 0.34161 3.2629   0.02 *
A1             1   0.4409 0.10256 2.9387   0.04 *
Management:A1  3   0.5892 0.13705 1.3090   0.20  
Residual      12   1.8004 0.41878                
Total         19   4.2990 1.00000                
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

And with version 2.6.10

> data(dune)
> data(dune.env)
> adonis2(dune ~ Management*A1, data=dune.env, permutations=99)
Permutation test for adonis under reduced model
Permutation: free
Number of permutations: 99

adonis2(formula = dune ~ Management * A1, data = dune.env, permutations = 99)
         Df SumOfSqs      R2      F Pr(>F)   
Model     7   2.4987 0.58122 2.3792   0.01 **
Residual 12   1.8004 0.41878                 
Total    19   4.2990 1.00000                 
---
Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1

Why were the explanatory variables removed from the aov table? Is there a way to get back a table including them?

[gavinsimpson]

Please check the NEWS for the version you are using where this change is documented and the solution is given (you want to add by = "terms"). This makes the anova() method for adonis2() compatible with the methods for other constrained ordinations. (This is also covered in several Issues/Discussions here already.)

[PierreQuevreux]

Thanks a lot!

[jarioksa]

First reported as item #694.