README.md

In wkmor1/msmod: Fit Multispecies Models

msmod

msmod is an R package for building statistical models of occupancy and abundance for more than one species at a time. So far, msmod includes the models described in Pollock etal, 2012 and Pollock etal 2014

Install external dependencies

First install the software package JAGS, appropriate for your platform, from this link

Install R package dependencies

Now run the following:

for (i in c('arm', 'devtools', 'dplyr', 'ggplot2', 'lme4', 'MASS', 'mclust', 'R2jags', 'rstan'))
if(!require(i, char = TRUE)) install.packages(i)

Install msmod

Install msmod from this repository using devtools.

devtools::install_github('wkmor1/msmod')

Fitting multispecies models

Fitting a multispecies trait model as in Pollock etal, 2012

library(msmod)
head(eucs)

##   plot logit_rock   ln_mrvbf ln_prec_yr ln_cv_temp ln_rad_d21 sand  loam
## 1    1  0.3846743 -0.7874579   6.665684   4.955827   8.719972 TRUE FALSE
## 2    2  1.0330150 -0.7874579   6.665684   4.955827   8.719972 TRUE FALSE
## 3    3  0.8001193 -0.7874579   6.659294   4.955827   8.087321 TRUE FALSE
## 4    4 -0.3846743 -0.7874579   6.656727   4.955827   7.400395 TRUE FALSE
## 5    5 -1.6034499 -0.7874579   6.648985   4.955827   7.209399 TRUE FALSE
## 6    6  1.2950457 -0.7874579   6.734592   4.955827   8.520882 TRUE FALSE
## 1   FALSE       0.89  35.57    35     ALA
## 2   FALSE       0.89  35.57    35     ALA
## 3   FALSE       0.89  35.57    35     ALA
## 4   FALSE       0.89  35.57    35     ALA
## 5   FALSE       0.89  35.57    35     ALA
## 6   FALSE       0.89  35.57    35     ALA

msm_glmer <- msm(y = "present", sites = "plot", x = "logit_rock",
  species = "species", traits = "ln_sla", data = eucs, type = "mstm",
  method = "glmer")

summary(msm_glmer)

## Generalized linear mixed model fit by maximum likelihood (Laplace Approximation) [
## glmerMod]
##  Family: binomial  ( logit )
## Formula: present ~ logit_rock + logit_rock:ln_sla + (1 + logit_rock | species)
##    Data: data
## Control: control
## 
##      AIC      BIC   logLik deviance df.resid 
##   4661.8   4704.6  -2324.9   4649.8     9154 
## 
## Scaled residuals: 
##     Min      1Q  Median      3Q     Max 
## -0.8182 -0.3615 -0.1997 -0.0736 15.3220 
## 
## Random effects:
##  Groups  Name        Variance Std.Dev. Corr 
##  species (Intercept) 2.0924   1.4465        
##          logit_rock  0.6376   0.7985   -0.13
## Number of obs: 9160, groups:  species, 20
## 
## Fixed effects:
##                   Estimate Std. Error z value Pr(>|z|)    
## (Intercept)        -3.2951     0.3367  -9.786  < 2e-16 ***
## logit_rock         -0.7262     0.2724  -2.666  0.00768 ** 
## logit_rock:ln_sla  -4.6150     0.5749  -8.027 9.97e-16 ***
## ---
## Signif. codes:  0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
## 
## Correlation of Fixed Effects:
##             (Intr) lgt_rc
## logit_rock  -0.006       
## lgt_rck:ln_  0.157  0.340

coef_plot(msm_glmer)

te_plot(msm_glmer, "logit_rock", "ln_sla")

wkmor1/msmod documentation built on May 4, 2019, 7:34 a.m.