Order of conditional variables changes outcome partial RDA

[britteheijink]

Hi all,

I'm running a partial RDA with three conditioning variables. I noticed that the resulting RDA object is different based on the order of the conditioning variabales (i.e. eigenvalues and inertia values are different). Is it not possible to run a partial RDA with multiple conditional variables or am I doing something wrong/is this expected?

A simplified example of my code is:

rda(x ~a + b +c, Condition(d + e + f), data = z, scale = TRUE)

I'm working in vegan 2.6.4 and R version 4.2.2

Thank you for your help!

[jarioksa]

It won’t. We need a reproducible example. BTW, your syntax is wrong. It should be y ~ x + Condition(z), data =… (NB the + sign).

[britteheijink]

I did use the + sign in my actual code, but found another syntax error in the actual code. I used Condition (d, e ,f) vs Condition (d +e +f). Problem solved. Thank you for your time and sorry to bother you for this kind of mistake!