In paul-vdb/DFO-master-sample: Western Canada Marine Master Sample

Data Input:

Load in the example shape files that cover the different marine protected areas (MPAs). We will do some data manipulation to ensure that join depth bands (0-20, 20-100, 100+) into strata along with the types of

## Should have a couple of libraries to start:
library(data.table)
library(ggplot2)
library(leaflet)
library(sp)

## Load the data from the package example:
data(NWT_Lakes)
daring.lks <- NWT_Lakes[SubBasin2 == "Daring"]

## Make it into a shape file with the correct projection:
pts.daring <- SpatialPointsDataFrame(SpatialPoints(cbind(daring.lks$easting, daring.lks$northing), 
    proj4string = CRS("+init=epsg:26912")), data = daring.lks)

Site Selection

Select Sites for each stratum using the NWT Freshwater Master Sample.

High CE:

Start with high cumulative effects. The CEV is the combined metric for natural, anthropogenic and vulnerability.

P1.h <- c(1,0)      #Stratum Permutation
P2.h <- c(1,0,2)    #Stratum Permutation
B.h <- 12   #Number of partitions
n.h <- 10   #Sample Size

## High subset:
pts.high <- pts.daring[pts.daring$CEV == "High", ]
## Get the Sites selected from the Master Sample:
smp.high <- MasterSample(pts.high, B = B.h, P1 = P1.h, P2 = P2.h, n = n.h)

## Make sure we track that this is the high stratum for later.
smp.high$Stratum <- "High"

Medium CE:

Now do it for Medium CE.

P1.m <- c(0,1)      #Stratum Permutation
P2.m <-  c(2,1,0)   #Stratum Permutation
B.m <- 12   #Number of partitions
n.m <- 10   #Sample Size

## High subset:
pts.med <- pts.daring[pts.daring$CEV == "Medium", ]
## Get the Sites selected from the Master Sample:
smp.med <- MasterSample(pts.med, B = B.m, P1 = P1.m, P2 = P2.m, n = n.m)

## Make sure we track that this is the high stratum for later.
smp.med$Stratum <- "Medium"

Low CE:

Now do it for Low CE.

P1.l <- c(0,1)      #Stratum Permutation
P2.l <-  c(0, 1, 2) #Stratum Permutation
B.l <- 12   #Number of partitions
n.l <- 10   #Sample Size

## High subset:
pts.low <- pts.daring[pts.daring$CEV == "Low", ]
# Get the Sites selected from the Master Sample:
smp.low <- MasterSample(pts.low, B = B.l, P1 = P1.l, P2 = P2.l, n = n.l)

## Make sure we track that this is the high stratum for later.
smp.low$Stratum <- "Low"

Visualization of the Sample:

We can make a nice looking plot in leaflet to visualize what we've done. Then we'll export the sample as a shape file or a csv.

pilot.pts <- rbind(smp.low, smp.med, smp.high)
pts.wgs <- spTransform(pilot.pts, CRS("+proj=longlat"))
pts.wgs.df <- data.frame(pts.wgs)
cols <- colorFactor(c("red", "orange", "green"), levels = c("High", "Medium", "Low"))
pts.wgs.df$col <- factor(pts.wgs.df$Stratum, levels =  c("High", "Medium", "Low"), labels = c("red", "orange", "green"))

## Visualization as an HTML map:
leaflet(pts.wgs.df) %>% addProviderTiles(providers$Esri.WorldImagery) %>%  
            addAwesomeMarkers(lng =~ coords.x1, lat =~ coords.x2, 
                icon=awesomeIcons(
                  icon = 'ios-close',
                  iconColor = 'black',
                  library = 'ion',
                  markerColor =~ col), popup = ~paste0("Stratum: ", Stratum, "SampleIndex: ", 
                        SampleIndex, "<br> HIPOrder: ", HIPOrder, 
                        "<br> Master Sample Index: ", MasterSampleIndex))

## Data export:
results.pts <- data.frame(pilot.pts)
write.csv(results.pts, "PilotPoints.csv", rownames = FALSE)

## Or as a shape file for GIS:
writeOGR(pilot.pts, "ExampleOutput", "PilotPoints", "ESRI Shapefile")