---
title: "Metaanalysis_Example"
author: "Bernd Gruber & Carlos González-Orozco"
date: "Wednesday, June 17, 2015"
output:
  word_document: default
  pdf_document:
    number_sections: yes
    toc: yes
  html_document:
    toc: yes
---

#Example code to run a metaanalysis to assess biodiversity using phylogenetic methods across multiple taxonomic groups (see González-Orozco et al.) 



The script presents all steps necessary to do the analysis of González-Orozco et al. (submitted) and consists of the following steps. Please note the full data set cannot be published at this state, due to copyright issues from external data providers, we will run the analysis on a subset of the data used in the manuscript. Therefore the results differ from the ones in the manuscript. The analysis comprises the following steps:

	1. LOAD DATA
	2. DIVERSITY AND ENDEMISM ANALYSES OF SINGLE and MULTIPLE TAXONOMIC GROUPS
	3. FUZZY CLUSTER ANALYSES OF INDIVIDUAL TAXONOMIC GROUPS
	4. PHYLOGENETIC CLUSTER ANALYSES OF INDIVIDUAL TAXONOMIC GROUPS

To run the example script you need to have the following files and folders:

* Metaanalysis_example.Rmd
* helper_functions.r
* data folder (which contains the subfolders grid, mck and shape)

Simply execute the code provided in the gray boxes in the pdf file step by step or easier load the rmd file into rstudio and press the Knit button and select the document type you want to produce.

![How to process the RMD file](./data/image/knitbutton.png)



# Load data

The data sets are gridded presence/absence data in Biodiverse format ([Lafan et al. 2010](https://github.com/shawnlaffan/biodiverse/wiki/PublicationsList)). For a description how to run biodiverse please refer to the biodiverse home page <http://purl.org/biodiverse>. As mentioned above the provided data comprises a subset of the original data set only.

Laffan, S.W., Lubarsky, E. & Rosauer, D.F. (2010) Biodiverse, a tool for the spatial analysis of biological and related diversity. Ecography. Vol 33, 643-647 (Version 1.0)


```{r setup, warning=FALSE, message=FALSE}
library(raster)
library(SDMTools)
library(rgdal)
#set the working directory 
setwd("d:/bernd/projects/crn/meta-analysis/")
#load some support functions
source("helper_functions.R")
```

# DIVERSITY AND ENDEMISM ANALYSES OF SINGLE AND MULTIPLE TAXONOMIC GROUPS (Figure 2 and 5)

The code below creates maps of biodiversity patterns via the function complots. Basically we use the value at each coordinate of a certain column (e.g. Species richness="ENDC_RI", please refer to Biodiverse manuals or check the headings in the example data set, which options are available) and standardise it if specified.
A combined plot summing all values of each species at each grid cell(e.g.combined species richness is also produced).

You can provide the following arguments to complots: 

* ident = a string specifying which column in the data set is to be used e.g. "ENDC_RI"
* scale = a switch if the values should be scaled between 0 and 1
* ncols = number of colours to be used 
* save = a switch if you want to save the plot in the working directory
* subsamp = a threhold that can be used to leave out data below a threshold
* complete = a switch: values for grid cells are only calculated if all species provide data for this cell (=FALSE) or at least on species provide data (=TRUE)


```{r makemaps, warning=FALSE, message=FALSE}
#create an empty list to load the data of each species
spec <- list()
#Find files of species data (needs to have grind in the name)
fn <- list.files("./data/grid/",pattern="grid")
#Load the data and store it in spec
for (i in 1:length(fn))
{
spec[[i]] <- read.csv(paste0("./data/grid/",fn[i]))
names(spec)[i] <- strsplit(fn[i],"_")[[1]][1]
}
#combine the data set into a single table specdat
specdat <- spec[[1]]
for (i in 2:length(fn))
specdat <- merge(specdat, spec[[i]][,-c(2:3)], by="Element", all=T)
#load shapefile (EsRI format)
map <-  readOGR("./data/shape/mdb.shp", "mdb", verbose=FALSE)
analysis <- c("ENDC_RI", "ENDC_WE" )# ,"PD_P", "PE_WE_P" ,"P_ENDC_WE", "P_PD_P"
#run a loop over each specified biodiversty index
for (i in 1:length(analysis))
{
complots(ident = analysis[i], scale=T, ncols=100, save=T, subsamp=-1, complete=F)
}
```

# DIVERSITY AND ENDEMISM ANALYSES OF INDIVIDUAL TAXONOMIC GROUPS (Figure 3)

For information how to create the CANAPE plots and analyse diversity and endemism please refer to the methods described in Mishler et al. (2014).

Mishler, B.D., Knerr, N., González-Orozco, C.E., Thornhill, A.H., Laffan, S.W., Miller, J.T. (2014) Phylogenetic measures of biodiversity and neo- and paleo-endemism in Australian Acacia. Nature Communications, 5, 4473


# FUZZY CLUSTER ANALYSES OF INDIVIDUAL TAXONOMIC GROUPS (Figure 4)

The maps on biodiversity indices were used as input to the Map Comparison Kit (MCK; Visser & Nijs 2006) to apply a fuzzy clustering analysis to evaluate the similarity of these maps. The resulting dissimilarity values of the MCK were used within 
a cluster analysis.

Visser, H., de Nijs, T. (2006) The Map Comparison Kit. Environmental Modeling & Software, 21, 346-358.

```{r}

#Example using Phylogenetic Endemism
srp <- read.csv("./data/mck/pe_pairs_all.csv")
m <- matrix(NA,nrow=5, ncol=5)
nn <-c("acacias", "eucs", "fishes", "frogs", "plants") 

for (i in 1:10)
{
m[which(nn==srp$sp1[i]), which(nn==srp$sp2[i])] <- srp$pe[i]
m[which(nn==srp$sp2[i]), which(nn==srp$sp1[i])] <- srp$pe[i]
}
mm <- as.dist(1-m)
plot(hclust(mm ), labels=nn,ylim=c(0,1),main="Phylogenetic Endemism", sub="",xlab=""  )

```