While other shape file functions can accept multiple states, or get all states if the states value is not specififed, the places() function only accepts a single state and must not be null. If given a vector of more than 1 state, it uses the first state's FIPS code for each state.

shape.places <- places(state = datasets::state.abb, cb = T)
Using FIPS code '19' for state 'AL'Using FIPS code '19' for state 'AK'Using FIPS code '19' for state 'AZ'Using FIPS code '19' for state 'AR'Using FIPS code '19' for state 'CA'Using FIPS code '19' for state 'CO'Using FIPS code '19' for state 'CT'Using FIPS code '19' for state 'DE'Using FIPS code '19' for state 'FL'Using FIPS code '19' for state 'GA'Using FIPS code '19' for state 'HI'Using FIPS code '19' for state 'ID'Using FIPS code '19' for state 'IL'Using FIPS code '19' for state 'IN'Using FIPS code '19' for state 'IA'Using FIPS code '19' for state 'KS'Using FIPS code '19' for state 'KY'Using FIPS code '19' for state 'LA'Using FIPS code '19' for state 'ME'Using FIPS code '19' for state 'MD'Using FIPS code '19' for state 'MA'Using FIPS code '19' for state 'MI'Using FIPS code '19' for state 'MN'Using FIPS code '19' for state 'MS'Using FIPS code '19' for state 'MO'Using FIPS code '19' for state 'MT'Using FIPS code '19' for state 'NE'Using FIPS code '19' for state 'NV'Using FIPS code '19'... <truncated>
  |=================================================================================================================================| 100%
Warning messages:
1: In if (grepl("^[[:digit:]]+$", state)) { :
  the condition has length > 1 and only the first element will be used
2: In if (grepl("^[[:alpha:]]+", state)) { :
  the condition has length > 1 and only the first element will be used
3: In if (nchar(state) == 2 & state %in% fips_state_table$abb) { :
  the condition has length > 1 and only the first element will be used
4: In fips_state_table$abb == state :
  longer object length is not a multiple of shorter object length
5: In fips_state_table$abb == state :
  longer object length is not a multiple of shorter object length

On Windows (with options(tigris_use_cache = TRUE)) , this gives me an error:

Sys.setenv(TIGRIS_CACHE_DIR = "~/.tigris_cache")
ca_districts <- school_districts(state = "06")

Error in ogrInfo(dsn = dsn, layer = layer, encoding = encoding, use_iconv = use_iconv,  : 
  Cannot open data source

But this works fine:

Sys.setenv(TIGRIS_CACHE_DIR = path.expand("~/.tigris_cache"))
ca_districts <- school_districts(state = "06")

It looks like this happens because rgdal::readOGR does not expand the path (so I can recreate the same errors by trying readOGR("~/.tigris_cache", "tl_2016_06_unsd")). The problem only occurs when importing to class sp, so this works fine:

Sys.setenv(TIGRIS_CACHE_DIR = "~/.tigris_cache")
ca_districts <- school_districts(state = "06", class = "sf")

At the moment on my computer, GET requests by tigris are being rejected, meaning that functions will fail when tigris_use_cache == TRUE. download.file(), used otherwise, is working fine.

I've put in a note to the Census Bureau to see if there is a reason for this behavior - otherwise I'll need to modify the code to use download.file() instead.

Ref: http://stackoverflow.com/questions/38206785/running-tigris-package-from-terminal

#!/usr/bin/Rscript --vanilla

library(tigris)
shp <- tracts(state = 'New York', county = 'New York')
print(shp)

Run that via ./script.R and it doesn't work. Cut/paste into R or RStudio and it works.

Using 0.5.1 on macOS, R 3.4:

gapuma <-  pumas("Georgia")
## Using FIPS code '13' for state 'Georgia'
## trying URL 'http://www2.census.gov/geo/tiger/TIGER%s/PUMA/tl_%s_%s_puma10.zip2015201513'
## Error in download.file(url, tiger_file, mode = "wb") : 
##   cannot open URL 'http://www2.census.gov/geo/tiger/TIGER%s/PUMA/tl_%s_%s_puma10.zip2015201513'
## In addition: Warning message:
##   In download.file(url, tiger_file, mode = "wb") :
##   cannot open URL 'https://www2.census.gov/geo/tiger/TIGER%s/PUMA/tl_%s_%s_puma10.zip2015201513': HTTP status was '400 Bad Request'

Traveling to the URL confirms that the page isn't there. I did find the file I needed at https://www2.census.gov/geo/tiger/TIGER2012/PUMA/tl_2012_13_puma10.zip. It seems as though the url constructor isn't building the string correctly.

It works, however, if I turn cb = TRUE

I get a big frowny face when I try to read the tutorial as directed on the R Project site for this package:

http://rpubs.com/walkerke/tigris01

The Census Bureau also provides "Cartographic Boundary Shapefiles" that are much smaller in size than the TIGER/Line shapefiles: https://www.census.gov/geo/maps-data/data/tiger-cart-boundary.html. This would be a good alternative for use cases that don't need all the detail in TIGER; I should look into making this an option.

I am using your package to calculate county road density for every U.S. county. I have processed 2328 counties but the function breaks on one specific county in South Dakota. South Dakota counties before and after appear to work. Code and session info below.

Great package!

st.val="South Dakota"
fips.val="113"

#Read Roads for state and county
in.roads <- roads( st.val, fips.val, year=2015)
Using FIPS code '46' for state 'South Dakota'
Previous download failed. Re-download attempt 1 of 3...
Downloading: 150 kB
Previous download failed. Re-download attempt 2 of 3...
Downloading: 150 kB
Previous download failed. Re-download attempt 3 of 3...
Downloading: 150 kB
Error: Download failed; check your internet connection or the status of the Census Bureau website
at http://www2.census.gov/geo/tiger/.

sessionInfo(package="tigris")
R version 3.3.2 (2016-10-31)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 7 x64 (build 7601) Service Pack 1
attached base packages:
character(0)
other attached packages:
[1] tigris_0.5.3

The sf package (https://cran.rstudio.com/web/packages/sf/) is now on CRAN; consider adding support for this in tigris. Certainly users could just convert Spatial*DataFrames to sf objects, but tigris could also read in the Census shapefiles this way directly.

I haven't pushed it to my fork yet, but I'm thinking that we control the caching & refresh with two options:

• options(tigris_refresh=FALSE)
• options(tigris_use_cashe=TRUE)

users can set these right after they load the package but these wld be the .onLoad defaults.

If that's cool, I'll try to push the PR tonight/tomorrow.

Hi, I want to implement this, and think I know how.

Two quick questions:

1. Any problem with me introducing a new potential value for the "tigris" attribute? It would be "sdall" or similar.

2. Any issue with renaming (with an explicit call out maybe) the UNSDLEA, SCSDLEA, ELSDLEA columns to "SDLEA" if the user uses the "all" argument for school_district(type = ...)?

Example of what I'm trying to do, and an example of the sort of workaround I want to implement in school_district()

> ky_unified <- school_districts(state = 'ky', type = 'unified')
Using FIPS code '21' for state 'KY'
> ky_sec <- school_districts(state = 'ky', type = 'secondary')
Using FIPS code '21' for state 'KY'
> temp <- rbind_tigris(ky_unified, ky_sec)
Error: Objects must all be Spatial*DataFrame objects and all the same type of tigris object.
> attr(ky_unified, "tigris") <- "sdall"
> attr(ky_sec, "tigris") <- "sdall"
> temp <- rbind_tigris(ky_unified, ky_sec)
Error in match.names(clabs, names(xi)) : 
  names do not match previous names
> names(ky_unified)[2] <- "SDLEA"
> names(ky_sec)[2] <- "SDLEA"
> temp <- rbind_tigris(ky_unified, ky_sec)
> head(as.data.frame(temp))
STATEFP SDLEA   GEOID                              NAME LSAD LOGRADE HIGRADE MTFCC SDTYP FUNCSTAT      ALAND
4015f6da0fef11e799e7e75995afe6670      21 00030 2100030      Adair County School District   00      PK      12 G5420  <NA>        E 1049678094
4015f6da0fef11e799e7e75995afe6671      21 00070 2100070      Allen County School District   00      PK      12 G5420  <NA>        E  891830403
4015f6da0fef11e799e7e75995afe6672      21 00078 2100078   Fort Campbell Dependent Schools   00      KG      12 G5420     B        E  146020316
4015f6da0fef11e799e7e75995afe6673      21 00079 2100079       Fort Knox Dependent Schools   00      KG      12 G5420     B        E  439748749
4015f6da0fef11e799e7e75995afe6674      21 00081 2100081 Muhlenberg County School District   00      PK      12 G5420  <NA>        E 1209734539
4015f6da0fef11e799e7e75995afe6675      21 00120 2100120   Anderson County School District   00      PK      12 G5420  <NA>        E  522745701
AWATER    INTPTLAT     INTPTLON
4015f6da0fef11e799e7e75995afe6670 18430783 +37.1055589 -085.2813796
4015f6da0fef11e799e7e75995afe6671 19492346 +36.7507703 -086.1924580
4015f6da0fef11e799e7e75995afe6672   197739 +36.6776414 -087.7370236
4015f6da0fef11e799e7e75995afe6673  6974504 +37.9058675 -085.8884215
4015f6da0fef11e799e7e75995afe6674 31906703 +37.2138159 -087.1340919
4015f6da0fef11e799e7e75995afe6675  6311536 +38.0053963 -084.9864171

Thanks!

GDal 2.x introduced int64, which R doesn't support (as of yet).
rgdal::readOGR changes the way it interprets numeric fields with values outside int32 range ( outside -2^32 to +2^31). with Gdal 1.x as there was no int64 the fields were interpreted as Real and converted to numeric, but if Gdal 2.x, any numeric field beyond int32 is handled using the 'interger64' argument of readOGR.

The current default is 'allow.loss' which silently truncates anything outside int32 range. This means the ALAND, AWATER fields of most shapefiles have truncated data.

https://gist.github.com/bhaskarvk/686deca51e3e81cb785ed8f2430ac2fc

Proposed workaround:-
Add integer64='no.loss' to the readOGR call, and then convert the chars for these fields to numeric. I doubt any of these fields will have AREA > 2^53 (MAX double), so the char to numeric (double) conversion should not result in any loss of data.

rgdal will probably be changed to make 'no.loss' the default option at some point in time, but for now this is a pressing issue especially for anyone using the ALAND/AWATER or for that matter any numeric field with values beyond the int32 range .

Given the way that the data are prepared from the Census Bureau - and how I am handling coordinate systems for the historic, generalized data - there are some slight datum inconsistencies that are causing problems when a user wants to combine data from different years in tidycensus. Take a look at this.

Maybe I'm just a little dense but I've been looking at the docs for ~10 minutes now and come up empty. Can I somehow set the cache directory?

Attempts to download tract boundaries for years prior to 2011 are returning the following error:

library(tigris)
#> 
#> Attaching package: 'tigris'
#> The following object is masked from 'package:graphics':
#> 
#>     plot
tigris::tracts(state = "53", county = "033", year = 2010)
#> Warning in unzip(file_loc, exdir = cache_dir, overwrite = TRUE): error 1 in
#> extracting from zip file
#> Error in ogrInfo(dsn = dsn, layer = layer, encoding = encoding, use_iconv = use_iconv, : Cannot open layer

An FYI for tigris users: the Census TIGER/Line website has been under maintenance for a couple days, so you currently aren't able to download data. If you are getting an error from tigris, this is why. I'll close the issue when the site is back up.

Hi, your package is really useful. I've been using it to download census blocks by county and state. Often, however, I need only a few blocks and downloading an entire county can be time consuming. Is there a way to download specific blocks or block groups instead. Thanks so much.

so, the resultant shapefile from a call to rails() results in a 353.3MB SpatialLinesDataFrame. I've got a killer system and even it struggles to do anything useful with it. Since even the d/l take a bit, perhaps we should take advantage of the know user-wait on first grab to split it into chunks based on state? Perhaps even add state and county params to the function like the others to give the illusion of similar operation to the user. We'd have to add a message saying what's going on to the user for the initial processing and this one would almost require (IMO) mandatory caching vs the optional one now.

Hi. This is a great package that I use quite a bit -- thanks! Found what I'm guessing is an easy to correct issue:

library(tigris)
cd115 <- congressional_districts(year=2016, refresh=TRUE)

yields the error:

downloaded 0 bytes

Error in download.file(url, tiger_file, mode = "wb") : 
  cannot download all files
In addition: Warning message:
In download.file(url, tiger_file, mode = "wb") :
  URL 'https://www2.census.gov/geo/tiger/TIGER2016/CD/tl_2016_us_cd114.zip': status was '404 Not Found'

...because the url should reference cd_115, not cd_114.

This url is valid: https://www2.census.gov/geo/tiger/TIGER2016/CD/tl_2016_us_cd115.zip

For reference:

> sessionInfo()
R version 3.3.1 (2016-06-21)
Platform: x86_64-apple-darwin13.4.0 (64-bit)
Running under: OS X 10.11.6 (El Capitan)

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] tigris_0.5.3 cbdb_0.1.0   RJDBC_0.2-5  rJava_0.9-8  DBI_0.6-13  

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.11.2  lattice_0.20-33 assertthat_0.1  dplyr_0.5.0     sf_0.5-4        rappdirs_0.3.1 
 [7] grid_3.3.1      R6_2.2.1        magrittr_1.5    units_0.4-6     httr_1.2.1      stringi_1.1.1  
[13] uuid_0.1-2      sp_1.2-3        rgdal_1.1-10    tools_3.3.1     stringr_1.2.0   udunits2_0.13  
[19] foreign_0.8-66  maptools_0.8-39 rgeos_0.3-19    tibble_1.2     

• Add all remaining geographies
• Give option to download very-generalized cartographic boundary files when applicable
• Remove warning message (if possible) associated with download progress bar
• Fully document the functions with descriptions from the Census Bureau
• Update the package for the imminent 2015 TIGER/Line shapefiles release
• Finish vignettes. I'm thinking three: one for basic usage, one for visualization, and one for spatial analysis.

The 1990 and 2000 cartographic boundary files give a separate row to each part of multipart polygons (e.g. counties with offshore islands). I was keeping this as-is to maintain fidelity to the source files (but correcting in tidycensus) but I am now cleaning this up to keep things consistent throughout the package and because it is likely confusing to users.

This all needs to be corrected before the 0.6 CRAN release.

This is the example code when I type ?states

library(tigris)
library(leaflet)

states <- states(detailed=FALSE)

leaflet(states) %>%
  addProviderTiles("CartoDB.Positron") %>%
  addPolygons(fillColor = "white",
              color = "black",
              weight = 0.5) %>%
  setView(-98.5795, 39.8282, zoom=3)

When I run it I get the following errors:

> states <- states(detailed=FALSE)
Error in load_tiger(url, tigris_type = "state", ...) : 
  unused argument (detailed = FALSE)
> 
> leaflet(states) %>%
+     addProviderTiles("CartoDB.Positron") %>%
+     addPolygons(fillColor = "white",
+                 color = "black",
+                 weight = 0.5) %>%
+     setView(-98.5795, 39.8282, zoom=3)
Error in polygonData.default(data) : 
  Don't know how to get path data from object of class function

Great package! I noticed that tigris does not consistently return sf objects when I use the class = 'sf' option. I think it happens when I use a vector of three digit FIPS codes for the county argument. Here is an easy way to reproduce the problem:

test1 <- tigris::blocks(state = "36", county = c("061", "047"), year = 2010, class = "sf")
class(test1)
# ERROR: test1 is class `SpatialPolygonsDataFrame` and not `sf`
test2 <- tigris::blocks(state = "36", county = "061", year = 2010, class = "sf")
class(test2)
## test2 is class `sf` so it works well here

Great work on the package, really good idea. But I was wondering, do any of these functions pull the cb_2013_us_county_ files? I looked through the code and couldn’t find it anywhere. I use that one a lot. Is it in there and I just missed it?

When installing tigris from github using devtools::install_github('walkerke/tigris') I get the following error:

Error in dyn.load(file, DLLpath = DLLpath, ...) : unable to load shared object '/home/user_name/R/x86_64-pc-linux-gnu-library/3.4/sf/libs/sf.so': /home/user_name/R/x86_64-pc-linux-gnu-library/3.4/sf/libs/sf.so: undefined symbol: _ZN17GDALDriverManager14GetDriverCountEv ERROR: lazy loading failed for package ‘tigris’ Error: Command failed (1)

Here is my R sessioninfo:

R version 3.4.2 (2017-09-28)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: Ubuntu 17.10

Matrix products: default
BLAS: /usr/lib/x86_64-linux-gnu/openblas/libblas.so.3
LAPACK: /usr/lib/x86_64-linux-gnu/libopenblasp-r0.2.20.so

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C               LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8     LC_MONETARY=en_US.UTF-8   
 [6] LC_MESSAGES=en_US.UTF-8    LC_PAPER=en_US.UTF-8       LC_NAME=C                  LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] devtools_1.12.0 lazyeval_0.2.0  rgdal_1.2-15    sp_1.2-5       

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.13    knitr_1.15.1    magrittr_1.5    maptools_0.9-2  units_0.4-6     rappdirs_0.3.1  lattice_0.20-35 R6_2.2.2       
 [9] udunits2_0.13   httr_1.3.1      tools_3.4.2     grid_3.4.2      e1071_1.6-8     DBI_0.6-1       git2r_0.18.0    withr_2.0.0    
[17] rgeos_0.3-26    class_7.3-14    digest_0.6.12   yaml_2.1.14     curl_3.0        memoise_1.1.0   compiler_3.4.2  classInt_0.1-24
[25] foreign_0.8-69

Thanks.

The build fails on travis because rgdal cannot be installed. Investigating this.

I'm using the cb argument to download a generalized and detailed version of a county file. When I look at the object sizes they seem to be about what you'd expect with a bigger object for the detailed layer. BUT.. when I plot them the "generalized" version seems a lot more detailed. Any thoughts about this?

library(tigris)
library(sp)
dtl <- counties(state="Wisconsin", cb=FALSE)
object.size(dtl) #1938576 bytes 
gen <- counties(state="Wisconsin", cb=TRUE)
object.size(gen) #500248 bytes

par(mfrow=c(1,2))
plot(dtl, main="Detailed?")
plot(gen, main="Generalized?")

Here is the plot

When rendering maps of the 50 states it is common to render AK and HI as insets, rather than in their "real" place. I searched but did not see an option to get maps from the tigris package in this way. Can you please confirm whether or not this is an option or not?

If it's not an option, can you please let me know whether it's on your roadmap?

Thanks for the great package.

I can't seem to download from AL, AK and TX. They all share the following error. I've tried both CRAN version: 0.3.3 and Dev version: 0.3.9.9000.

> library(tigris)
> state_legislative_districts("TX", house = 'lower')
Using FIPS code '48' for state 'TX'
Error in ogrInfo(dsn = dsn, layer = layer, encoding = encoding, use_iconv = use_iconv,  : 
  Cannot open layer
In addition: Warning message:
In unzip(file_loc, exdir = cache_dir, overwrite = TRUE) :
  error 1 in extracting from zip file

we (@bi-sdal ) had a huge issue with downloading data in tigris.
Mainly because files were not being downloaded properly.
A few things we had to to do to get the tigris working (we tested the package with va <- states('51'))

1. must set options(tigris_use_cache = TRUE), otherwise we get the following error:

> va <- states('51')
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
Error in ogrInfo(dsn = dsn, layer = layer, encoding = encoding, use_iconv = use_iconv,  : 
  Cannot open data source
In addition: Warning message:
In unzip(tiger_file, exdir = tmp) : error 1 in extracting from zip file

2. install.packages(c("curl", "httr"))
   • we tried to install the github version of the package, but was running in to SSL CA cert problems (see r-lib/devtools#1079)
   • if we do not install these 2 packages we get the following error:

> options(tigris_use_cache = TRUE)
> va <- states('51')              
Previous download failed.  Re-download attempt 1 of 3...
Previous download failed.  Re-download attempt 2 of 3...
Previous download failed.  Re-download attempt 3 of 3...
Error: Download failed; check your internet connection or the status of the Census Bureau website
                 at http://www2.census.gov/geo/tiger/.

Here's the sessionInfo() without installing curl and httr

> sessionInfo()
R version 3.4.0 (2017-04-21)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: CentOS Linux 7 (Core)

Matrix products: default
BLAS: /usr/lib64/microsoft-r/3.4/lib64/R/lib/libRblas.so
LAPACK: /usr/lib64/microsoft-r/3.4/lib64/R/lib/libRlapack.so

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C              
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8    
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                 
 [9] LC_ADDRESS=C               LC_TELEPHONE=C            
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
[1] tigris_0.5.3         RevoUtilsMath_10.0.0

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.12     bindr_0.1        magrittr_1.5     maptools_0.9-2  
 [5] units_0.4-5      rappdirs_0.3.1   uuid_0.1-2       lattice_0.20-35 
 [9] R6_2.2.0         rlang_0.1.1      stringr_1.2.0    udunits2_0.13   
[13] httr_1.2.1       dplyr_0.7.1      tools_3.4.0      rgdal_1.2-8     
[17] grid_3.4.0       DBI_0.7          rgeos_0.3-23     assertthat_0.2.0
[21] tibble_1.3.3     sf_0.5-2         bindrcpp_0.2     curl_2.6        
[25] glue_1.1.1       sp_1.2-5         stringi_1.1.5    compiler_3.4.0  
[29] RevoUtils_10.0.4 foreign_0.8-67   pkgconfig_2.0.1 

cc: @dads2busy

I am new to R - so if this is stupid, my apologies. A recurring error happens when the app attempts to download the share files:

Error in ogrInfo(dsn = dsn, layer = layer, encoding = encoding, use_iconv = use_iconv, :
Cannot open data source
In addition: Warning message:
In unzip(file_loc, exdir = cache_dir, overwrite = TRUE) :
error 1 in extracting from zip file
This occurs when I run your example code:

library(tigris)
library(magrittr)
ua <- urban_areas(cb = TRUE)
plot(ua)

I really hope I can resolve this. Your package would be tremendously useful to me!

I see that you're importing sp but I'm surprised that the following won't work:

library(tigris)
x<-counties(c("RI", "DE"))
plot(x)
#Error in as.double(y) : 
#  cannot coerce type 'S4' to vector of type 'double'

Instead it seems you need to explicitly load sp to get the sp plotting methods.

library(tigris)
library(sp)
x<-counties(c("RI", "DE"))
plot(x)

I think that this could be confusing for relative beginners, perhaps load the sp plotting methods in the package?

Take a look at this.

I want to permanently set the obj class that tigris returns as sf, so I added the option to my .Rprofile options: options(... , tigris_class = 'sf').

This works great:

> getOption('tigris_class')
[1] "sf"

But when I attach the package to use it, my already set tigris_* options get overwritten which causes me the additional step to reset the option in the script:

> library(tigris)
As of version 0.5.1, tigris does not cache downloaded data by default. To enable caching of data, set `options(tigris_use_cache = TRUE)` in your R script or .Rprofile.
Attaching package: ‘tigris’
The following object is masked from ‘package:graphics’:
    plot

> getOption('tigris_class')
[1] "sp"

> options(tigris_class = 'sf')
> getOption('tigris_class')
[1] "sf"

Thanks!

Session info ------------------
 setting  value                       
 version  R version 3.3.3 (2017-03-06)
 system   x86_64, mingw32             
 ui       RStudio (1.0.143)           
 language (EN)                        
 collate  English_United States.1252  
 tz       America/Chicago             
 date     2017-06-29                  

Packages ------------------------
 package    * version  date       source        
...
 sf           0.5-1    2017-06-23 CRAN (R 3.3.3)
 tidycensus   0.1.2    2017-06-20 CRAN (R 3.3.3)
 tigris     * 0.5.3    2017-05-26 CRAN (R 3.3.3)
...

How difficult would it be to add this year? Need it since that's when the last decennial Census was done.

Census seems to have moved the files referenced in the package so that when one tries to use the "congressional_districts" function, the following error message is returned:

co_cd <- congressional_districts("Colorado", cb = TRUE, resolution = '20m')
trying URL 'http://www2.census.gov/geo/tiger/GENZColorado/shp/cb_Colorado_us_cd114_20m.zip'
Error in download.file(url, tiger_file, mode = "wb") :
cannot open URL 'http://www2.census.gov/geo/tiger/GENZColorado/shp/cb_Colorado_us_cd114_20m.zip'
In addition: Warning message:
In download.file(url, tiger_file, mode = "wb") :
cannot open URL 'http://www2.census.gov/geo/tiger/GENZColorado/shp/cb_Colorado_us_cd114_20m.zip': HTTP status was '404 Not Found'

@hrbrmstr I wrote a validate_county function based on your validate_state function, and it is working after a few tests; take a look when you get the chance and let me know if there is anything you think should be cleaned up

I just installed R 3.4.1 and tigris 0.5.3 on Windows 10. When I try to run any download functions (states(), counties()) I get an error that the URL cannot be opened. For example

> spdfCounty = counties(state = "NY", refresh = TRUE)
trying URL 'http://www2.census.gov/geo/tiger/TIGER2015/COUNTY/tl_2015_us_county.zip'
Error in download.file(url, tiger_file, mode = "wb") : 
  cannot open URL 'http://www2.census.gov/geo/tiger/TIGER2015/COUNTY/tl_2015_us_county.zip'
In addition: Warning message:
In download.file(url, tiger_file, mode = "wb") :
  InternetOpenUrl failed: 'A redirect request will change a non-secure to a secure connection'

This had been working for me in R 3.3.1 (but a few months ago, just upgraded the version on this computer). It is also working on a Linux computer (R 3.4.1, tigris 0.5.3).

FWIW, I've confirmed that the URL works in a browser. I've tried calling download.file directly, and that fails. However, if I change the URL to ftp://ftp.census.gov/... that works!

This seems to be a very similar issue to #14 and #16, but my solution is opposite!

In those issues people were unable to write to a cache folder, so setting tigris_use_cache=FALSE solved their issues, but for me FALSE is the default setting and isn't working, but setting tigris_use_cache=TRUE does work for me. Moreover, I'm importing tigris with a package I'm distributing, and I'd like to have both options, cache or not, working, or be able to advise users of my package how to configure their systems so that the portions of my package that use tigris will work.

> options(tigris_use_cache=FALSE) # default on 0.5.3
> me <- counties("Maine", cb = TRUE)
  % Total    % Received % Xferd  Average Speed   Time    Time     Time  Current
                                 Dload  Upload   Total   Spent    Left  Speed
  0     0    0     0    0     0      0      0 --:--:-- --:--:-- --:--:--     0
Error in ogrInfo(dsn = dsn, layer = layer, encoding = encoding, use_iconv = use_iconv,  :
  Cannot open data source
In addition: Warning message:
In unzip(tiger_file, exdir = tmp) : error 1 in extracting from zip file
> options(tigris_use_cache=TRUE)
> me <- counties("Maine", cb=TRUE)
Using FIPS code '23' for state 'Maine'

my sessionInfo()

R version 3.4.3 (2017-11-30)
Platform: x86_64-pc-linux-gnu (64-bit)
Running under: CentOS Linux 7 (Core)

Matrix products: default
BLAS: /home/cmarmstrong/miniconda2/lib/R/lib/libRblas.so
LAPACK: /home/cmarmstrong/miniconda2/lib/R/lib/libRlapack.so

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8
 [7] LC_PAPER=en_US.UTF-8       LC_NAME=C
 [9] LC_ADDRESS=C               LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

other attached packages:
[1] tigris_0.5.3         rlandscan_0.0.0.9000 RevoUtils_10.0.8
[4] RevoUtilsMath_10.0.1

loaded via a namespace (and not attached):
 [1] Rcpp_0.12.14     bindr_0.1        magrittr_1.5     maptools_0.9-2
 [5] units_0.4-6      rappdirs_0.3.1   uuid_0.1-2       lattice_0.20-35
 [9] R6_2.2.2         rlang_0.1.6      stringr_1.2.0    udunits2_0.13
[13] httr_1.3.1       dplyr_0.7.4      tools_3.4.3      rgdal_1.2-16
[17] grid_3.4.3       e1071_1.6-8      DBI_0.7          rgeos_0.3-26
[21] class_7.3-14     assertthat_0.2.0 tibble_1.4.1     sf_0.5-5
[25] bindrcpp_0.2     glue_1.2.0       sp_1.2-5         stringi_1.1.6
[29] pillar_1.0.1     compiler_3.4.3   classInt_0.1-24  foreign_0.8-69
[33] pkgconfig_2.0.1

I don't know much about R and temporary directories, but I checked tempdir() and the folder it returns, and it belongs to me so I'd think my R session could write to it also.

It seems that the only year arguments that succeed in the blocks() function are 2014 and 2015.

> wa_blk <- tigris::blocks(state = 53,county = 33,year = 2012) 
Downloading: 140 kB     
error 1 in extracting from zip fileError in ogrInfo(dsn = dsn, layer = layer, encoding = encoding, use_iconv = use_iconv,  : 
  Cannot open layer

t <- area_water("VA","515")

Throws an error. FIPS code "515" is for "Bedford City", which is returned by list_counties("VA"). All other VA counties returned by list_counties work.

one more for the potential PR. Adding a tigris attr that will contain the type TIGER object being returned by the various tigris functions. That way, we can make an rbind_tigris function that will let folks retrieve, say, all the area water shapefiles for a state then rbind them together, auto-creating unique IDs per county. It'll also help for (perhaps) plotting and summary functions.

Possibly related to/dupe of #16:

I ran

tigris::tracts(state='CA')

then aborted the download before it could complete. Upon re-executing the command I got

> tigris::tracts(state='CA')
Using FIPS code '06' for state 'CA'
Error in ogrInfo(dsn = dsn, layer = layer, encoding = encoding, use_iconv = use_iconv,  : 
  Cannot open layer
In addition: Warning message:
In unzip(file_loc, exdir = cache_dir, overwrite = TRUE) :
  error 1 in extracting from zip file

I'm assuming this is because it's trying to unzip the partially downloaded zip from the first try instead of downloading the zip anew? Forcing a re-download using refresh=TRUE as suggested in #16 fixes things. But perhaps tigris should be smarter about recognizing such corrupted files?

I use tigris with my class. The zctas functions works wonderfully. But now we want to work with commuting zones.

1. I do not see commuting zones as available in tigris. Am I missing something?

2. I see commuting zone information in your US Boundaries package. However, I can't even load the data there into R. Any advice?

Thanks!

As I have time I'm adding in the remaining Census geographies that I haven't gotten to yet, as @hrbrmstr you've prepared a great template for what they should look like at this point. I'll push them to GH as I go.

Hey All : big fan of the tigris package but having trouble with block data for certain states:

tigris::blocks( state= '06', county = '099', year = 2015)

immediately results in :
Error in ogrInfo(dsn = dsn, layer = layer, encoding = encoding, use_iconv = use_iconv, :
Cannot open layer
In addition: Warning message:
In unzip(file_loc, exdir = cache_dir, overwrite = TRUE) :
error 1 in extracting from zip file

many other state and county blocks queries return data as expected. any ideas ?

i am using R version 3.3.1 (2016-06-21) and tigris_0.3.9.9000

HI, I work in an environment where we can't get rgdal 2.0+ version on our servers. Is there a way to install tigris without the sf package which only works with rgdal 2.0+?

• add longer-form docs to a Description fields for various functions. this is just a matter of cutting/pasting salient text from the census site
• add a @seealso with a URL to the salient Census URL for said functions

I just wanted to start off by saying thank you for creating this package! It's so incredibly useful!
I also raised this issue on the tmap Github page here: https://github.com/mtennekes/tmap/issues/165.

I'll start from the beginning. I want to create my own interactive maps of the United States using data I've obtained from various sources. After searching around, tmap seemed like a good solution for me. After poking around for some tutorials, I found this one for creating interactive choropleths: https://github.com/mtennekes/tmap/blob/master/demo/USChoropleth/US_choropleth.R.

The demo worked! I now wanted to add tigris to my workflow. So I created the following R script which pretty much follows the above demo code with only slight changes adding in tigris:

library(readxl)
library(grid)
library(tmap)
library(tmaptools)
library(tigris)
#options(tigris_class = "sf")

# function to obtain Food Environment Atlas data (2014)
get_food_envir_data <- function() {
  dir <- tempdir()
  download.file("https://www.ers.usda.gov/webdocs/DataFiles/48731/February2014.xls?v=41688", destfile = file.path(dir, "DataDownload.xls"), mode="wb")
  read_excel(file.path(dir, "DataDownload.xls"), sheet = "HEALTH")
}

# obtain Food Environment Atlas data
FEA <- get_food_envir_data()

# obtain US county shape
US_c <- counties()

# create FIPS variable
US_c$FIPS <- paste0(US_c$STATEFP, US_c$COUNTYFP)

# append data to shape
US_c <- append_data(US_c, FEA, key.shp = "FIPS", key.data = "FIPS", ignore.duplicates = TRUE)

# interactive choropleth
ttm()
qtm(US_c, fill = "PCT_OBESE_ADULTS10")

However, this causes my RStudio application to crash. The reason being my RAM is totally overloaded as R quickly uses up all of 8 GB of RAM.

I tried to figure out why it was when I used tigris I was crashing R. I quickly found out that the shape files obtained in the tmap demo were nearly 10x smaller than the shape files obtained through tigris. Yet, when you draw a static map using the two different shape files, the maps of the U.S. counties seem to be the same!

So my question is, am I doing something wrong when I call the U.S. counties? Why are my shape files so big? Any help would be greatly appreciated.