How to create a `geofacet` grid from the Guerry map of France in 1830?

Question

I'm trying to create a grid for the geofacet package based on the map of France from 1830 in the Guerry package.

The base map, Guerry::gfrance85 is a SpatialPolygonsDataFrame ofthe 85 departments (excluding Corsica) at that time

I have two problems:

• Creating a reasonable grid from the centroids of the departments in the map. The one I created below has too many holes.
• Getting the geofacet package to recognize this [But I'll save that for a separate query]

Below is my working MRE. I chose to use a 17 x 17 grid because any smaller gives duplicate cells for the departments.

library(sp)
library(geofacet)
library(ggplot2)
library(dplyr)
#library(tidyr)
library(Guerry)


# scale a variable into n integer bins
bin <- function(x, n=10){
  1 + floor(n * (x - min(x)) / (max(x) - min(x)))
}

data(gfrance85, package = "Guerry")

dept         <- data.frame(gfrance85)[,"dept"]   
dep.names    <- data.frame(gfrance85)[,"Department"]
region.names <- data.frame(gfrance85)[,"Region"]


# extract department centroids
xy <- coordinates(gfrance85) |>
  as.data.frame() |>
  setNames(c("x", "y")) 

# make the grid data.frame
nbin <- 17   # size of grid
gfrance85_grid <- data.frame(code = dept,
                             name = as.character(dep.names),
                             row = bin(xy$x, nbin),
                             col = bin(xy$y, nbin))
head(gfrance85_grid)

My grid data set looks like this:

  code         name row col
1    1          Ain  15   8
2    2        Aisne  12  16
3    3       Allier  11   9
4    4 Basses-Alpes  16   4
5    5 Hautes-Alpes  16   5
6    7      Ardeche  13   5

When I plot this, as shown below, it shows the first problem: There are too many holes in the grid. How can I process this data set to close some of the holes (without having duplicate cells)?

gfrance85_grid |>
  mutate(label = paste(code, "\n", name)) |>
  ggplot(aes(x = row, y = col)) +
  geom_tile(color = "black",
            fill = scales::alpha("blue", .3)) +
  geom_text(aes(label = label), size = 2) +
  coord_equal()