tidyr error with oc_forward_df

[lsong07]

I just tried only geocoding the first 5000 records of my dataframe in R and get this error, seeming like to be caused with tidyr when oc_forward_df automatically tries to clean the API responses:

result <- opencage::oc_forward_df(batch, placename = full_address)
Retrieving results from OpenCage [|] 100% ETA: 0s
Error in list_unchop():
! Can't combine x[[1]]$oc_building and x[[1280]]$oc_building .
Run rlang::last_trace() to see where the error occurred.
rlang::last_trace()
<error/vctrs_error_ptype2>
Error in list_unchop():
! Can't combine x[[1]]$oc_building and x[[1280]]$oc_building .

---

Backtrace:
▆

1. ├─opencage::oc_forward_df(batch, placename = full_address)
2. └─opencage:::oc_forward_df.data.frame(batch, placename = full_address)
3. ├─tidyr::unnest(results_nest, .data$op, names_repair = "unique")
4. └─tidyr:::unnest.data.frame(results_nest, .data$op, names_repair = "unique")

5. └─tidyr::unchop(...)
   

6.   └─tidyr:::df_unchop(...)
   

7.     └─vctrs::list_unchop(col, ptype = col_ptype)
   

Run rlang::last_trace(drop = FALSE) to see 9 hidden frames. > rlang::last_trace(drop = FALSE)
<error/vctrs_error_ptype2>
Error in list_unchop():
! Can't combine x[[1]]$oc_building and x[[1280]]$oc_building .

Backtrace:
▆

1. ├─opencage::oc_forward_df(batch, placename = full_address)
2. ├─opencage:::oc_forward_df.data.frame(batch, placename = full_address)
3. │ ├─tidyr::unnest(results_nest, .data$op, names_repair = "unique")
4. │ └─tidyr:::unnest.data.frame(results_nest, .data$op, names_repair = "unique")
5. │ └─tidyr::unchop(...)
6. │ └─tidyr:::df_unchop(...)
7. │ └─vctrs::list_unchop(col, ptype = col_ptype)
8. └─vctrs (local) ()
9. └─vctrs::vec_default_ptype2(...)

10. ├─base::withRestarts(...)
    

11. │ └─base (local) withOneRestart(expr, restarts[[1L]])
    

12. │   └─base (local) doWithOneRestart(return(expr), restart)
    

13. └─vctrs::stop_incompatible_type(...)
    

14.   └─vctrs:::stop_incompatible(...)
    

15.     └─vctrs:::stop_vctrs(...)
    

16.       └─rlang::abort(message, class = c(class, "vctrs_error"), ..., call = call)
    

-->

Session information

R version 4.4.1 (2024-06-14 ucrt)
Platform: x86_64-w64-mingw32/x64
Running under: Windows 11 x64 (build 22631)

Matrix products: default

locale:
[1] LC_COLLATE=English_United States.utf8 LC_CTYPE=English_United States.utf8 LC_MONETARY=English_United States.utf8 LC_NUMERIC=C LC_TIME=English_United States.utf8

time zone: America/Los_Angeles
tzcode source: internal

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

other attached packages:
[1] furrr_0.3.1 future_1.34.0 ggmap_4.0.0 ggplot2_3.5.1 opencage_0.2.2 purrr_1.0.2 dplyr_1.1.4 httr_1.4.7 tidygeocoder_1.0.5

loaded via a namespace (and not attached):
[1] utf8_1.2.4 generics_0.1.3 tidyr_1.3.1 bitops_1.0-9 jpeg_0.1-10 stringi_1.8.4 httpcode_0.3.0 hms_1.1.3 listenv_0.9.1 digest_0.6.37 magrittr_2.0.3 grid_4.4.1 fastmap_1.2.0 jsonlite_1.8.9 plyr_1.8.9
[16] progress_1.2.3 crul_1.5.0 urltools_1.7.3 fansi_1.0.6 scales_1.3.0 codetools_0.2-20 cli_3.6.3 crayon_1.5.3 rlang_1.1.4 triebeard_0.4.1 parallelly_1.38.0 munsell_0.5.1 withr_3.0.1 cachem_1.1.0 tools_4.4.1
[31] parallel_4.4.1 memoise_2.0.1 colorspace_2.1-1 ratelimitr_0.4.1 globals_0.16.3 curl_6.0.0 assertthat_0.2.1 vctrs_0.6.5 R6_2.5.1 png_0.1-8 lifecycle_1.0.4 stringr_1.5.1 pkgconfig_2.0.3 pillar_1.9.0 gtable_0.3.5
[46] glue_1.8.0 Rcpp_1.0.13 tibble_3.2.1 tidyselect_1.2.1 compiler_4.4.1 prettyunits_1.2.0

[freyfogle]

HI, Ed from OpenCage here.

I know nothing about R, but is your code assuming there will always be a building value?
That is not the case.

[dpprdan]

@lsong07 We've seen errors like this before when the OpenCage API returns results for the same field (here building) with different types (usually some string for some results and a number for others). So this could be the case here as well.

Could you please provide the full_address of the rows that failed (apparently rows 1 and 1280), so we can check?