AMRO-Interactions.Rnw

\documentclass{article}
\usepackage{pdflscape}
\usepackage{longtable}
\usepackage{parskip}

\title{Follow-up analysis American Robin (\textsc{AMRO}) nest success along trails\\ (with interactions)}
\author{Max A Henschell}
\begin{document}
% \SweaveOpts{concordance=TRUE}
% % \SweaveOpts{concordance=TRUE}
% \SweaveOpts{width=10, height = 8}

\maketitle

This document is follow-up analysis following comments from \textit{Landscape and Urban Planning}. The reviewers suggested that combining data from Marty (herein \textsc{MJP}) and Max (herein \textsc{MAH}) was not appropriate.\\
Below are the results from separating the analysis of the two studies. I used \texttt{BMA} with no interactions (not shown) to guide the selection of variables in the final models,% exlcuding variables with $(P{\neq}0) < 0.2$, 
plus incorporated seemingly ecologically meaninful interactions into the final full model.
\newpage
% Figure ~\ref{fig:MJPCorr} shows the correlation between measured variables for the \textsc{MJP} portion of the study.\\
% Figure ~\ref{fig:MAHCorr} shows the correlation between measured variables for the \textsc{MAH} portion of the study.\\


%Correlation plots for individual studies are presented first, followed by the results of individual and combined Bayesian model averaging (\textsc{BMA}).

<<include=FALSE, echo=FALSE, results='hide'>>=
source("AMRO_InteractionModels_MJP.R")
source("AMRO_InteractionModels_MAH.R")
source("AMRO_InteractionModels_Combined.R")
lapply(packages.list, require, character.only=T)
@

%\section*{Variable Correlations}

%\begin{landscape}
%' \begin{figure}[!htbp]
%' \centering
%' <<echo=FALSE, results='hide'>>=
%' ####################################
%' #
%' #Correlation for MJP
%' #
%' ####################################
%' corrplot.mixed(cor(AMRO.MJP.nl[,c(11:23)]))
%' 
%' @
%' \caption{Correlation of potential explanatory variables for \textsc{MJP AMRO} nests}
%' \label{fig:MJPCorr}
%' \end{figure}
%' \begin{center}
%' \end{center}
%' %\end{landscape}
%' 
%' %\begin{landscape}
%' \begin{figure}[!htbp]
%' \centering
%' <<echo=FALSE, results='hide'>>=
%' ####################################
%' #
%' #Correlation MAH
%' #
%' ####################################
%' corrplot.mixed(cor(AMRO.MAH.nl[,c(11:23)]))
%' @
%' \caption{Correlation of potential explanatory variables  for \textsc{MAH AMRO} nests}
%' \label{fig:MAHCorr}
%' \end{figure}
%' \begin{center}
%' \end{center}
%\end{landscape}
% variables:\\
% Forest1km: proportion forest cover within 1 km of nest -> all studies\\
% Road1km: km of roads within 1 km of nest -> all studies\\
% Homesteads1km : number of homesteads within 1 km of nest -> all studies\\
% RoadDist: distance to closest road -> all studies\\
% W2: trail width at 1.3 m from ground -> all studies\\
% GPH: groups per hour -> MAH,MJP\\
% TPH: trees pre hectare around nest -> MAH\\
% BA\_ha: basal area per hectare around nest -> MAH,MJP\\
% TYPE: trail type: constructed, old road -> all studies\\
% PLOT: trail name -> all studies\\
% HEIGHT: nest height -> all studies\\
% DIST: distance from nest to trail -> all studies\\
% ORIENT: Nest orientation from trail: 90 = opposite side of tree from trail\\
% OrdDate: Ordinal date from Jan 1 - 140 -> all studies\\
% STUDY: study\\
% YEAR: year of nest\\
% NEST: nest ID\\


\begin{figure}[!htbp]
<<echo=FALSE, results='hide'>>=
imageplot.bma(AMRO.MJP.BMA, color = c("blue", "red", "white"), order = "mds")
@
\caption{Variable selection for \texttt{BMA} from \textsc{MJP} data.
\textcolor{red}{Red} indicates a negative association with nest success, \textcolor{blue}{blue} indicates positive.}
\label{fig:MJPBMAimage}
\end{figure}


\begin{figure}[!htbp]
<<echo=FALSE, results='hide'>>=
imageplot.bma(AMRO.MAH.BMA, color = c("blue", "red", "white"), order = "mds")
@
\caption{Variable selection for \texttt{BMA} from \textsc{MAH} data.
\textcolor{red}{Red} indicates a negative association with nest success, \textcolor{blue}{blue} indicates positive.}
\label{fig:MAHBMAimage}
\end{figure}

%\begin{landscape}
<<echo=FALSE, results= 'asis'>>=
xtable(MJP.BMA.df[-c((nrow(MJP.BMA.df)-3):nrow(MJP.BMA.df)),c(1:3)], caption = "Tabular results from \\texttt{BMA} for \\textsc{MJP} data")
xtable(MAH.BMA.df[-c((nrow(MAH.BMA.df)-3):nrow(MAH.BMA.df)),c(1:3)], caption = "Tabular results from \\texttt{BMA} for \\textsc{MAH} data")
#xtable(BMA.df[-c((nrow(BMA.df)-3):nrow(BMA.df)),c(1:4)], caption = "Tabular results from \\texttt{BMA} for \\textsc{MJP+MAH} data")
@
%\end{landscape}

\begin{figure}[!htbp]
<<echo=FALSE, results='hide'>>=
imageplot.bma(AMRO.BMA, color = c("blue", "red", "white"), order = "mds")
@
\caption{Variable selection for \texttt{BMA} from \textsc{MAH+MJP} data.
\textcolor{red}{Red} indicates a negative association with nest success, \textcolor{blue}{blue} indicates positive.}
\label{fig:MAHBMAimage}
\end{figure}

%\begin{landscape}
<<echo=FALSE, results= 'asis'>>=
#xtable(MJP.BMA.df[-c((nrow(MJP.BMA.df)-3):nrow(MJP.BMA.df)),c(1:4)], caption = "Tabular results from \\texttt{BMA} for \\textsc{MJP} data")
#xtable(MAH.BMA.df[-c((nrow(MAH.BMA.df)-3):nrow(MAH.BMA.df)),c(1:4)], caption = "Tabular results from \\texttt{BMA} for \\textsc{MAH} data")
xtable(BMA.df[-c((nrow(BMA.df)-3):nrow(BMA.df)),c(1:3)], caption = "Tabular results from \\texttt{BMA} for \\textsc{MJP+MAH} data")
@
%\end{landscape}

\end{document}