mooreryan/divnet-rs

Diversity estimates not plausible

Closed this issue · 2 comments

ARetter90 commented

Hi,

When computing Shannon and Inverse Simpson diversities for my samples from the divnet-rs output, samples that have a very low observed richness have shannon diversities of 9 and higher - how can this be? This is biologically not logical if i only have say 5 species in that sample and such a high shannon index and inverse simspn numbers of >1000. Do you have any suggestion on how to deal with this?

Alice

mooreryan commented

Hi Alice, thanks for your report.

It's hard for me to say what's going on without seeing how you ran all the commands and your input data.

Would you be able to share your input data, the exact command line command you ran, and the R code in which you calculated the diversity measures? If so I can try to reproduce the issue.

Thanks!

mooreryan commented

I see that some members of the Willis lab have responded to this issue on the DivNet R page (adw96/DivNet#126).

I'm closing the issue here to avoid duplicated work between the two groups.

Feel free to reopen if it turns out to be an issue specific to divnet-rs, and not the DivNet method in general.