maps - Crop rasterLayer with SpatialPolygonDataFrame leaves out part of the target region R -

library(raster) library(ncdf4) library(rgdal) 

i trying crop raster layer based on polygon: sample data (~45mb) can found here sample data illustration:

dat<- raster('data.nc')# make subset  dat class       : rasterlayer  dimensions  : 824, 935, 770440  (nrow, ncol, ncell) resolution  : 10000, 10000  (x, y) extent      : -5000, 9345000, -5000, 8235000  (xmin, xmax, ymin, ymax) coord. ref. : na  data source : c:\users\data.nc  names       : total.precipitation  zvar        : total.precipitation  

as coord. ref = na, assign polar stereographic projection based on information website of data provider.

projj=crs("+proj=stere +lat_0=90 +lat_ts=90 +lon_0=-100 +k=0.994 +x_0=2000000 +y_0=2000000 +datum=wgs84 +units=m +no_defs") proj4string(dat) <- projj     #rx <- projectraster(from=dat, crs=wgs84.p4s) 

then transform bbox shapefile projj

    shgrid <- sptransform(prairie.boundaries, projj)# transform dat spat reference     plot(dat)     plot(shgrid,add=t)# gives 

cropping

    cr <- crop(dat, extent(shgrid))      plot(cr)     fr <- rasterize(shgrid, cr)      dat1<- mask(x=cr, mask=fr)     levelplot(dat1) 

output:

the cropped image shows part of canada want of canada.

what did miss within code? projection?

besides prefer axes labels in latlon coordinates.

any suggestions?