Visible Data - 2022-09-01_ggplot2-breaks-geom-sf-calls-in-3-3-6

This is a draft blogpost.

As detailed here https://github.com/tidyverse/ggplot2/issues/4883 the 3.3.6 version of {ggplot2} introduces a new linewidth aesthetic. For ALL geoms except for geom_sf() and geom_pointrange() this causes no issues.

But in choropleth built with {ggplot2} where size = n was used to decrease the weight of polygon lines will no longer be effective. I tweeted about it here: https://twitter.com/charliejhadley/status/1562800738816106496?s=20&t=ixvEsEfT7lAgFfmJb3yTQg

I think that some static code analysis could help with this!

Identify the line numbers of all geom_sf() calls with size as a direct argument [ie not inside aes]
Optionally allow users to specify a range for the size value, predicting that the size being smaller than 1 is a useful hueretistic for identifying problematic code.

I started doing a bit of research:

library(tidyverse)

── Attaching packages ─────────────────────────────────────── tidyverse 1.3.2 ──
✔ ggplot2 3.3.6      ✔ purrr   0.3.5 
✔ tibble  3.1.8      ✔ dplyr   1.0.10
✔ tidyr   1.2.1      ✔ stringr 1.4.1 
✔ readr   2.1.3      ✔ forcats 0.5.2 
── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag()    masks stats::lag()

library(rnaturalearthdata)
library(sf)

Linking to GEOS 3.11.0, GDAL 3.5.3, PROJ 9.1.0; sf_use_s2() is TRUE

world_sf <- countries110 %>% 
  st_as_sf()

ggplot() +
  geom_sf(data = world_sf,
          size = 0.1)

ggplot() +
  geom_sf(data = world_sf,
          fill = "green",
          size = 0.4)

This parses out an expression - but doesn’t give me line numbers

parse(here::here("posts", "2022-09-01_ggplot2-breaks-geom-sf-calls-in-3-3-6", "code-to-analyse.R")) %>% 
  keep(is.language)  %>% 
  keep(~grepl(", geom_sf", toString(.x)))

expression(ggplot() + geom_sf(data = world_sf, size = 0.1), ggplot() + 
    geom_sf(data = world_sf, fill = "green", size = 0.4))

getParseData() is a really useful function that gives LOTS of data. But I’m not sure how to partse it yey

parse_data <- parse(here::here("posts", "2022-09-01_ggplot2-breaks-geom-sf-calls-in-3-3-6", "code-to-analyse.R")) %>% 
  getParseData() %>% 
  rownames_to_column() %>% 
  rename(parent_id = rowname) %>% 
  mutate(parent_id = as.integer(parent_id)) %>% 
  as_tibble()


parse_data %>% 
  relocate(parent_id, parent, token, text) %>% 
  filter(text == "geom_sf")

This is probably quite close to what I need https://stackoverflow.com/a/47189529

library(tidyverse)
library(rnaturalearthdata)
library(sf)
library(rlang)
world_sf <- countries110 %>% 
  st_as_sf()


fenv <- new.env()

evaluated_geom_sf <- parse(here::here("posts", "2022-09-01_ggplot2-breaks-geom-sf-calls-in-3-3-6", "code-to-analyse.R")) %>% 
  keep(is.language)  %>% 
  keep(~grepl(", geom_sf", toString(.x))) %>% 
  map(expression, envir=fenv) 

parse(here::here("posts", "2022-09-01_ggplot2-breaks-geom-sf-calls-in-3-3-6", "code-to-analyse.R")) %>% 
  keep(is.language)  %>% 
  keep(~grepl(", geom_sf", toString(.x))) %>% 
  map(expr_interp) %>% 
  map(call_args)

evaluated_geom_sf %>% 
  map(quote)

rlang::call_name(evaluated_geom_sf[1])

  map_df(~{
    params <- list(names(formals(.x)))
    bdy <- deparse(body(.x))
    bdy <- bdy[length(bdy)-1]
    data_frame(target = trimws(bdy), params = params)
  }) %>% 
  mutate(fname = ls(fenv))

fenv <- new.env()

evaluated_fns <- parse(here::here("posts", "2022-09-01_ggplot2-breaks-geom-sf-calls-in-3-3-6", "function-code.R")) %>% 
  keep(is.language)  %>% 
  keep(~grepl(", function", toString(.x))) %>% 
  map(eval, envir=fenv) %>% 
  map_df(~{
    params <- list(names(formals(.x)))
    bdy <- deparse(body(.x))
    bdy <- bdy[length(bdy)-1]
    data_frame(target = trimws(bdy), params = params)
  }) %>% 
  mutate(fname = ls(fenv))

evaluated_fns %>% 
  unnest(params)

Reuse

https://creativecommons.org/licenses/by/4.0/

Citation

BibTeX citation:

@online{hadley2022,
  author = {Charlie Hadley},
  title = {2022-09-01\_ggplot2-Breaks-Geom-Sf-Calls-in-3-3-6},
  date = {2022-09-01},
  url = {https://visibledata.co.uk/posts/2022-09-01_ggplot2-breaks-geom-sf-calls-in-3-3-6},
  langid = {en}
}

For attribution, please cite this work as:

Charlie Hadley. 2022. “2022-09-01_ggplot2-Breaks-Geom-Sf-Calls-in-3-3-6.” September 1, 2022. https://visibledata.co.uk/posts/2022-09-01_ggplot2-breaks-geom-sf-calls-in-3-3-6.