Extracting an xarray based netcdf file to use in R

xarray_to_r.r

    require(ncdf4)  
    require(ncdf4.helpers)
    require(data.table)
    
    ## Get the name of the value vars in the nc file
    get_nc_value_name <- function(nc_file) {
      
      ## Get names
      nc_obj <- nc_open(nc_file)
      name<-names(nc_obj$var)
      
      ## Close file
      nc_close(nc_obj)
      
      ## Return the name
      return(name)
      
    }
    
    ## Once we have the name of the variable we want to extract, we pass it onto this function to return the full dataset
    xarray_nc_to_R <- function(nc_file, dimname, start=NA, count=NA, df_return = T) {
      
      ## Open the file and show the attribuets
      ncin <- nc_open(nc_file)
      print(ncin)
      
      ## Get the full array, using the variable name we want
      Rarray <- ncvar_get(ncin, dimname, start = start, count = count, collapse_degen=F)
      
      ## Get the fillvalue info
      fillvalue <- ncatt_get(ncin,dimname,"_FillValue")
      
      ## Get the dimension names in the right order
      array_dim <- ncdf4.helpers::nc.get.dim.names(ncin, dimname)
      
      
      ## Close the file
      nc_close(ncin)
      
      ## Get all of the dimension information in the order specified
      array_dim_list <- list()
      for(i in array_dim) {
        array_dim_list[[i]] <- ncin$dim[[i]]$vals
      }
      
      ## Fill in NaNs with NA
      Rarray[Rarray==fillvalue$value] <- NA
      
      
      ## Assign the dimension labels to the R array
      for(i in 1:length(array_dim_list)) {
        dimnames(Rarray)[[i]] <- array_dim_list[[i]]  
      }
      
      ## Attach the dimension name to the array
      names(attributes(Rarray)$dimnames) <- array_dim
      
      if(df_return) {
        return(data.frame(reshape2::melt(Rarray)))  
      } else {
        return(Rarray)  
      }
      
      
      
    }
    

  
    
    ### Testing with Population data from FBD
    xarray_data <- "REDACTED"
    dimname = get_nc_value_name(xarray_data)
    
    ## We'll extract the 'population' column and convert to a DT
    ## NOTE: We are pulling in the full xarray. If you wanted to slice, then you'd have
    ## to define the start and count variables... TBV
    
    pop_data <- data.table(xarray_nc_to_R(nc_file = pop_file_2, dimname))
    print(pop_data)