Tutorial:LEA won't run lfmm sequentially and independently on individual environmental variables when d set to 0
Entering edit mode
Last seen 4.1 years ago
Harvard University, Cambridge, Massachu…

according to the package doc, when running lfmm in LEA, the def for -d is

"An integer corresponding to the fit of lfmm model with the d-th variable only
from environment.file. By default (if NULL and all is FALSE), fit lfmm with
each variable from environment.file sequentially and independently."

however, if I run it in the following fashion:

project = lfmm("oneperrad_trachylepis_RAD.geno", "taff_lea.env", iteration = 10000, burnin = 5000,d = 0,missing.data = TRUE, CPU = 8,K = 5, repetitions = 5, project = "new")

LEA exits without errors after finishing with the first variable. the logging indicates it is analyzing the first variable

as the default behavior is all=FALSE, I'm not sure why it isn't starting for the next variable. Incidentally, I'm running variable 2 right now using d=2, and it is chugging along just fine.

LEA lfmm Tutorial • 1.0k views
Entering edit mode

it seems that for LEA, 0 != NULL, but somehow translates into 1. Omitting the -d arg altogether results in lfmm iterating all of the environmental variables. go figure.

Entering edit mode
Last seen 4.2 years ago

Indeed, the LEA documentation and LEA behaviour is not clear about this subject !

Let me clarify:

  • If you do not specify a "-d" value, it will fit lfmm with each variable from environment.file sequentially and independently.
  • If you specify a d value below 0 (d=0, d=-1, ...), it will set d=1. LEA should at least raise a warning (or even an error about it).
  • If you specify a d value strictly upper 0 (d=1, d=2, ...), it will fit the model with the d-th variable only.

Thanks for pointing it out 



Login before adding your answer.

Traffic: 552 users visited in the last hour
Help About
Access RSS

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6