search

vegandevs / vegan

R package for community ecologists: popular ordination methods, ecological null models & diversity analysis
https://vegandevs.github.io/vegan/
GNU General Public License v2.0
446 stars 96 forks source link

CCA - varpart result discrepancy #418

Open francesco-ricci opened 3 years ago

francesco-ricci commented 3 years ago

Hello everyone, I am running variation partitioning (using cca instead of rda) and cca from the r package vegan. Both analyses are run on the same dataset, but I get different results in terms of the variation of the dataset that the two analyses explain. I wondered if anyone can help explain why despite both analyses relying on cca I get different results.

Thanks!

jarioksa commented 3 years ago

Do you mean that varpart is based on adjusted R2 instead of raw proportions of inertia? I think this is obvious both from the documentation and column headers in varpart. Please give more details specifying, how the results are inconsistent.

francesco-ricci commented 3 years ago

I run varpart (CCA) and CCA on the same X dataset.

varpart explains in total (counting each partition) 11% of the composition of X dataset.

CCA explains 17% of the composition of X dataset.

Since both test run on CCA, I wonder why the results are different.

Thanks

jarioksa commented 3 years ago

Adjusted R2 vs unadjusted R2.