CODEWITHSUNDEEP
rmapscensus
mmalloy
mmalloy

Reputation: 74

acs package in R: Cannot download dataset, error message is inscrutable

I am trying to use the acs package in R to download Census data for a basic map, but I am unable to download the data and I'm receiving a confusing error message.

My code is as follows: 

#Including all packages here in case this is somehow the issue

install.packages(c("choroplethr", "choroplethrMaps", "tidycensus", "tigris", "leaflet", "acs", "sf"))

library(choroplethr)
library(choroplethrMaps)
library(tidycensus)
library(tigris)
library(leaflet)
library(acs)
library(sf)
library(tidyverse)

api.key.install("my_api_key")

SD_geo <- geo.make(state="CA", county = 73, tract = "*", block.group = "*")

median_income <- acs.fetch(endyear = 2015, span = 5, geography = SD_geo, table.number = "B19013", col.names="pretty")

Everything appears to work until the final command, when I receive the following error message: 

trying URL 'http://web.mit.edu/eglenn/www/acs/acs-variables/acs_5yr_2015_var.xml.gz'
Content type 'application/xml' length 735879 bytes (718 KB)
downloaded 718 KB

Error in if (url.test["statusMessage"] != "OK") { : 
  missing value where TRUE/FALSE needed
In addition: Warning message:
In (function (endyear, span = 5, dataset = "acs", keyword, table.name,  :
  XML variable lookup tables for this request
  seem to be missing from ' https://api.census.gov/data/2015/acs5/variables.xml ';
  temporarily downloading and using archived copies instead;
  since this is *much* slower, recommend running
  acs.tables.install()

This is puzzling to me because 1) it appears like something is in fact being downloaded at first? and 2) 'Error in if (url.test["statusMessage"] != "OK") { : missing value where TRUE/FALSE needed' makes no sense to me. It doesn't align with any of the arguments in the function.

I have tried:

I'm going to just download the datafile as a CSV and read it into R for now, but I'd like to be able to perform this function from the script for future maps. Any information on what's going on here would be appreciated. I am running R version 3.3.2. Also, I'm new to using this package and the API. But I'm following the documentation and can't find evidence that I'm doing anything wrong.

Tutorial I am working off of: http://zevross.com/blog/2015/10/14/manipulating-and-mapping-us-census-data-in-r-using-the-acs-tigris-and-leaflet-packages-3/#get-the-tabular-data-acs

And documentation of the acs package: http://eglenn.scripts.mit.edu/citystate/wp-content/uploads/2013/02/wpid-working_with_acs_R2.pdf

Upvotes: 2

Views: 1024

Answers (3)

Ezra Haber Glenn
Ezra Haber Glenn

Reputation: 21

To follow up on Brandon's comment, version 2.1.1 of the package is now on CRAN, which should resolve this issue.

Upvotes: 2

Brandon Metzger
Brandon Metzger

Reputation: 11

I emailed Ezra Haber Glenn, the author of the package, about this as I was having the same issue. I received a response within 30 minutes and it was after midnight, which I thought was amazing. Long story short, the acs package version 2.1.0 is configured to work with the changes the Census Bureau is making to their API later this summer, and it is currently presenting some problems windows users in the mean time. Ezra is going to be releasing an update with a fix, but in the mean time I reverted back to version 2.0 and it works fine. I'm sure there are a few ways to do this, but I installed the devtools package and ran:

require(devtools)

install_version("acs", version = "2.0", repos = "http://cran.us.r-project.org")

Hope this helps anyone else having a similar issue.

Upvotes: 1

kwalkertcu
kwalkertcu

Reputation: 1061

Your code runs for me. My guess would be that the Census API was temporarily down.

As you loaded tidycensus and you'd like to do some mapping, you might also consider the following code: 

library(tidycensus)
census_api_key("your key here") # use `install = TRUE` to install the key
options(tigris_use_cache = TRUE) # optional - to cache the Census shapefile

median_income <- get_acs(geography = "block group", 
                         variables = "B19013_001", 
                         state = "CA", county = "San Diego", 
                         geometry = TRUE)

This will get you the data you need, along with feature geometry for mapping, as a tidy data frame.

Upvotes: 0

Related Questions