rmomal/emtree

Infers species direct association networks

emtree's Issues

0/1 matrix

Hi, Momal:
We want to use the inference network to do some further analysis, but we face some problems like: abundance ~ x1+x2+network(infer from abundance), so I am thinking whether we can infer the network from 0/1 matrix by using EMtree?

There's no ComparEMtree()

it seems there's no function "ComparEMtree" in this package

Post-treatments and warning messages

"data" is the species abundance matrix

##"site" is the environmant covariance matrix
##HU is a factor in site

fit PLN model

PLNMODEL<- PLN(data~site$HU)
Initialization...
Adjusting a PLN model with full covariance model
Post-treatments...
DONE!
Warning message:
In dpois(y, mu, log = TRUE) : non-integer x = 1.214953

Follow is my questions:

1. I am not sure whether the warning message is because of too limit data or too many zero in the data;
2. The model can still run and I can get the result of the inference network, so I am curious whether this result is robust;
3. I also want to know that if the result is not robust what can I do to infer species interaction network?

model dateError in terms.formula(object, data = data) : ExtractVars里的模型公式不对

compare_output<-ComparEMtree(counts, covar_matrix=covar, models=tested_models, m_names=models_names, Pt=0.15, S=3, maxIter=5,cond.tol=1e-8,cores=1)

model dateError in terms.formula(object, data = data) : ExtractVars里的模型公式不对