Is there a way to pass piped data through ggmap like you can with ggplot to the geoms?-CodePudding

While exploring with ggplot a common workflow is to do some data manipulation and then pipe directly into a ggplot(). When you do that, all of that manipulation flows nicely through the ggplot into the geoms with no variables needed. Like so:

data %>%
  filter(route %in% c('01','08','15')) %>%
  ggplot()   
  geom_sf()   
  geom_sf_text(aes(label=route))

If however you want to use a ggmap() for a nice background, then there doesn't appear to be a way to use the piped workflow. You have to save the manipulations to a variable first, which isn't a huge deal, but would love to know if there's a way to avoid it.

This doesn't work:

background = get_stamenmap(....) # somewhat irrelevant to the question

data %>%
  filter(route %in% c('01','08','15')) %>%
  ggmap(background)   
  geom_sf()   
  geom_sf_text(aes(label=route))

Thought maybe trying to force the data into ggmap() might help but it doesn't

background = get_stamenmap(....) # somewhat irrelevant to the question

data %>%
  filter(route %in% c('01','08','15')) %>%
  ggmap(background, data=.)   
  geom_sf()   
  geom_sf_text(aes(label=route))

Or maybe there's some other way to combine ggplot() and ggmap() to accomplish it? I know I can save off the manipulated data as a variable and then hard-code that into each geom_sf() layer, but it's just not as convenient and thought I might be missing something simple.

CodePudding user response：

The magrittr::%>% infix operator expects to pass the data as the first argument of the first expression in a -chain as you have here. Unfortunately, you want to pass it to one of the not-first expressions. You can use a {-block.

library(ggmap)
library(ggplot2)
library(dplyr) # for %>%, could do magrittr as well

### from ?get_stamenmap
bbox <- c(left = -97.1268, bottom = 31.536245, right = -97.099334, top = 31.559652)
background <- get_stamenmap(bbox, zoom = 14)

### from my brain
set.seed(42)
dat <- data.frame(x=runif(4, bbox[1], bbox[3]), y=runif(4, bbox[2], bbox[4]), lbl = sample(LETTERS, 4))
dat
#           x        y lbl
# 1 -97.10167 31.55127   Q
# 2 -97.10106 31.54840   O
# 3 -97.11894 31.55349   X
# 4 -97.10399 31.53940   G

dat %>% {
  ggmap(background)  
  geom_point(aes(x, y), data = .)  
  geom_text(aes(x, y, label = lbl), data = ., color = "red",
            hjust = 0, vjust = 0)
}