tidycensus_ri.R

library(tidycensus)
ri <- get_decennial(geography = "state", state = "RI",variables = "P001001", 
                    key = api_key, 
                    year = 2000)

#Returning this error...
#Using FIPS code '44' for state 'RI'
#Using FIPS code '44' for state 'RI'
#Error : lexical error: invalid char in json text.
#                                       error: unknown/unsupported geog
#                     (right here) ------^
#
#Error: is.character(x) is not TRUE          

I don't have that direct functionality exposed at the moment. What I recommend:

library(tidycensus)
library(tidyverse)

census_api_key("your key here")

ri <- get_decennial(geography = "state", variables = "P001001", year = 2000) %>%
  filter(NAME == "Rhode Island")

My philosophy in the package design has not been to fully expose the API, but rather to help users get enough of what they need and then do the rest within the tidyverse. If this functionality would be useful, however, I'll consider putting it on the roadmap. You can also look at @hrecht's censusapi package, which fully exposes the API: https://github.com/hrecht/censusapi.