Empty strings in character or factor variables

[vincentarelbundock]

Thanks for your work on this great package! I use it all the time and love it.

I’m wondering if it would be possible to handle empty strings more gracefully. For example, this CSV file includes empty strings in the sex column. If I read the file with read.csv(url,na.strings="") then tabular() works without issues. However, if I read the file with read.csv(url), then I get a cryptic error message. I feel that empty strings may be a valid category in some cases.

library(tables)

url <- 'https://vincentarelbundock.github.io/Rdatasets/csv/palmerpenguins/penguins.csv'

dat <- read.csv(url, na.strings = "")
tabular(Factor(sex) ~ bill_length_mm * mean * Arguments(na.rm = TRUE), data = dat)
#                       
#         bill_length_mm
#  sex    mean          
#  female 42.10         
#  male   45.85

dat <- read.csv(url)
tabular(Factor(sex) ~ bill_length_mm * mean * Arguments(na.rm = TRUE), data = dat)
# Error in as.name(catname): attempt to use zero-length variable name