How to convert Multipolygon column to sfc column? R

I´m having trouble converting coordinates data to sf format. I exported a csv file from QGIS with attributes of several political units with their respective coordinates. Because I'm working with polygons I exported the csv file from QGIS as WKT (MultiPolygon Z).

However, when I import the csv file into R, it takes the coordinate column as a character instead of sfg.Therefore, I use the st_as_sf function, but it returns the following error:

all_sf <- all2 %>% st_as_sf(wkt = "geometry", crs = 4326) 
OGR: Corrupt data Error: OGR error

(1000 rows more)

str(all2)'data.frame':  1058 obs. of  5 variables:
$ OBJECTID : chr  "13939" "13939" "13939" "13939" ...
$ Distrito : chr  "Alberti" "Alberti" "Alberti" "Alberti" ...
$ Año      : chr  "1996" "1997" "1998" "1999" ...
$ toneladas: num  0 0 0 0 0 0 0 0 0 0 ...
$ geometry : chr  "MULTIPOLYGON Z (((-60.32233429 -34.7621421809999 0,-60.2506828309999 -34.8179740909999 0,-60.2398338319999 -34."| truncated "MULTIPOLYGON Z (((-60.32233429 -34.7621421809999 0,-60.2506828309999 -34.8179740909999 0,-60.2398338319999 -34."| truncated "MULTIPOLYGON Z (((-60.32233429 -34.7621421809999 0,-60.2506828309999 -34.8179740909999 0,-60.2398338319999 -34."| truncated "MULTIPOLYGON Z (((-60.32233429 -34.7621421809999 0,-60.2506828309999 -34.8179740909999 0,-60.2398338319999 -34."| truncated ...

I believe there is either an issue with:

1. your data
2. legacy issues if you have previously installed packages that bind to GDAL, GEOS and PROJ such as sp

rather than sf misinterpreting your geometry values as strings. That's because this works:

library(rnaturalearth)
library(dplyr)
library(sf) # GEOS 3.11.2, GDAL 3.7.2, PROJ 9.3.0; sf_use_s2() is TRUE

packageVersion("rnaturalearth")
# [1] ‘1.0.1’
packageVersion("dplyr")
# [1] ‘1.1.4’
packageVersion("sf")
# [1] ‘1.0.15’

# Create sample csv file
sf_csv <- ne_countries(returnclass = 'sf') %>%
  mutate(sfid = 1:n()) %>%
  select(sovereignt, sfid) %>%
  st_write(., "sf_csv.csv",
           append = FALSE,
           layer_options = "GEOMETRY=AS_WKT")

# Writing layer `sf_csv' to data source `sf_csv.csv' using driver `CSV'
# options:        GEOMETRY=AS_WKT 
# Writing 177 features with 2 fields and geometry type Multi Polygon.

# Import sample csv
sf_csv <- read.csv("sf_csv.csv")

# Create sf object from sf_csv
all_sf <- sf_csv %>% st_as_sf(wkt = "WKT", crs = 4326) 
class(all_sf)
# [1] "sf"         "data.frame"

One solution is to see if exporting your QGIS data as a shapefile, then importing it into R solves your issue. First, in QGIS:

• Go to the Layers Panel on the left side
• Right-click the layer of interest
• Click on Save as
• Select Esri Shapefile
• Name file and select directory location
• Click OK

Next, import your shapefile data into R:

all_sf <- st_read("path/to/your/shapefile.shp") %>%
  st_set_crs(4326) # If your data do not have a CRS already

If this does not work, then at least you can confirm the error lies with your original QGIS data or your libraries and can start debugging from there.