Ggmap-geompoint, how to make grouping? - r

Question

Suppose I have this dataframe
latitude longitude category
42.39905 -72.93871 A
42.39905 -73.93871 B
43.37471 -73.36336 A
43.37471 -74.36336 B
44.28322 -74.31423 B
What I would like to do is to group the coordinates by its integer. Then for each group, I could create a bubble with a size function on the counts in a group.
The colour diverges from A to B, based on how many A than B. So far, I've been doing this,
map = get_map(location="jk",zoom=6,source="stamen")
#Plot the point
ggmap(map)+
geom_point(data=zipmap,
aes(x=round(longitude),y=round(latitude),colour=category))+
scale_color_brewer(type='div')
But as you would expect, the colour is not diverging, and the size of the bubble is not implemented. How could I achieve this? I can't use scale_x_continuous, as it already used somewhere in ggmap

Answer 1

Here is one direction to try.
dput(df)
structure(list(latitude = c(42.39905, 42.39905, 43.37471, 43.37471,
44.28322), longitude = c(-73, -74, -73, -74, -74), category = structure(c(1L,
2L, 1L, 2L, 2L), .Label = c("A", "B"), class = "factor"), latround = structure(c(1L,
1L, 2L, 2L, 3L), .Label = c("42", "43", "44"), class = "factor"),
longround = structure(c(2L, 1L, 2L, 1L, 1L), .Label = c("-74",
"-73"), class = "factor")), .Names = c("latitude", "longitude",
"category", "latround", "longround"), row.names = c(NA, -5L), class = "data.frame")
df$latround <- as.factor(round(df$latitude)) # round the coords
df$longround <- as.factor(round(df$longitude))
library(dplyr) # group by rounded coordinates and count the categories
df2 <- df %>% group_by(latround) %>% summarise(catnumber = n())
latround catnumber
1 42 2
2 43 2
3 44 1
library(ggmap)
From here you don't specify the location jk so I outlined an approach to plotting.
map <- get_map(location="jk",zoom=6,source="stamen")
#Plot the point
ggmap(map)+
geom_point(df2, aes(x=longround),y=latround), size = catnumber, colour=catnumber))+
scale_color_brewer(type='div') # more is needed in the ggmap code