behlpr: Functions to run my biodiversity analysis

#' I switched to processing in ArcGIS, but there are already fragmentation data from Catrin Westphal.
#' Calculate Fragmentation Data
#' 
#' Turned out this was hard to do in R, so I did it in QGIS instead. Here are the steps I took:
#' @details   Here are the steps I took:
#' \tabular{ll}{
#' Download CORINE 2006 (COR06) Data   \tab Accepted continent-wide fragmentation data \cr
#' Cut COR06 data to match each exploratory \tab work with smaller data files, done earlier in ArcGIS \cr
#' Reproject COR06 to match exploratories (UTM 32 for HAI & ALB, 33 for SCH). Chose cubic resampling - oh, I bet I know where I got the all forest cells from last time... \tab Better to have everything in same coord. system
#' Start with plot shapefiles (UTM 32 for HAI & ALB, 33 for SCH)  \tab I made these with ArcGIS based on the .kml files \cr
#' Set QGIS to do on-the-fly converstions \tab Helps get layers to line up \cr
#' Buffer plot shapefiles (250 m, 1000 m, 5000 m)      \tab \cr
#' }
#' 
unfinished.function = function(NOTHING){
  
  
  # Code out steps needed for CORINE Data
  
  # Load required packages
  library(raster)
  library(rgdal)
  
  # Load CORINE Data
  # Reference site: http://neondataskills.org/R/Raster-Data-In-R/
  corine = raster("C:/docs/beplants/datasets/CORINE_2006/g100_06.tif")
  
  # Load BE Exploratories plot locations
  plots.dir = "C:/docs/beplants/datasets/GIS/BExIS_data" #**# NOTE: Trailing slash crashes readOGR. I'd totally write a patch for this.
  plots = readOGR(dsn= plots.dir, layer="Grassland_EPs")
  
  # Project plots to UTM
  plots.utm = sp::spTransform(plots, CRS("+proj=utm +zone=32 ellps=WGS84"))
  
  
  #**# Trying to solve spatialpixels problem
  #?raster
  x = seq(479629,867282)
  y = seq(5301768, 5940455)
  vals = length(x) * length(y)
  z = runif(vals,1,10)
  
  raster(nrows = length(x), ncols = length(y), xmn = min(x), xmx = max(x), ymn = min(y), ymx = max(y), vals = z, crs = CRS("+proj=utm +zone=32 ellps=WGS84"))
  
  test = matrix(c(x,y,z), nrow = 3)
  raster(test)
  
  test1 = seq(1,100)
  test2 =
    
    test1 = rep(seq(1,100),2)
  test2 = sort(rep(seq(1,100),2))
  test0 = seq(1,200)
  test3 = matrix(c(test1,test2,test0), ncol = 3)
  
  test = raster(test3)
  test.sp = as(test, "SpatialPixels")
  
  # Get buffer distances around plots
  require(adehabitatMA)
  
  adehabitatMA::buffer(plots.utm, plots.utm, 50)
  #Error in adehabitatMA::buffer(plots, plots, 50) : 
  #  x should inherit the class SpatialPixels
  
  
  
  # Extract Corine data by buffer
  #**# This step will overlap with what Bastian is doing.
  
  #http://stackoverflow.com/questions/13982773/crop-for-spatialpolygonsdataframe
  
  # Reclassify Corine data to grassland, forest, & Other
  
  
  # Estimate % cover for grassland & forest
  
  
  # Estimate contagion for grassland habitat
  #**# Rodolphe had suggested doing this by edge density. I'm not exactly sure how to implement this in the raster context
  #**# but if I do, I can add it as an output to spatialdemography at the same time (highly desirable)
  
  # output data to be read in by R
  
  
  ## Helper functions (or is Rodolphe's edge density measure already coded in R?)
  #**# Start with internet search - try to avoid recreating the wheel. & talk w/Bastian
  
}