Errors in many `bake` methods when `newdata` has class `grouped_df`

[jrnold]

Many bake methods will either raise an error or return an empty dataset if newdata is a grouped data frame (classgrouped_df as returned by dplyr::group_by). The reproducible example below provides a few examples. This appears to affect any bake method which uses the pattern,

newdata <- cbind(newdata, ...)

The problem occurs because when newdata is "grouped_df" then this code will return a list rather than the expected data frame object.

Replacing instances of cbind with dplyr::bind_cols seems to solve the problem - it passes all tests and produces the desired results. I could submit a pull request, but I haven't fully grokked recipes, and I don't know if grouped data frames have other implications for recipes. Should groupings be removed? Have special treatment?

Minimal, runnable code:

library("tidyverse")
#> ── Attaching packages ────────────────────────────────────────────── tidyverse 1.2.1 ──
#> ✔ ggplot2 2.2.1          ✔ purrr   0.2.4     
#> ✔ tibble  1.4.2          ✔ dplyr   0.7.4.9000
#> ✔ tidyr   0.8.0          ✔ stringr 1.2.0     
#> ✔ readr   1.1.1          ✔ forcats 0.2.0
#> ── Conflicts ───────────────────────────────────────────────── tidyverse_conflicts() ──
#> ✖ dplyr::filter() masks stats::filter()
#> ✖ dplyr::lag()    masks stats::lag()
library("recipes")
#> Loading required package: broom
#> 
#> Attaching package: 'recipes'
#> The following object is masked from 'package:stringr':
#> 
#>     fixed
#> The following object is masked from 'package:stats':
#> 
#>     step
foo <- tibble(bar = letters, qux = runif(length(bar)), baz = rep(1:2, length = length(bar))) %>% 
  group_by(baz)

# error in step_dummy
rec <- recipe(~bar, data = foo) %>% step_dummy(bar)

trained_rec <- prep(rec, training = foo)

# error in merging indicators
try(bake(trained_rec, newdata = foo))
#> Warning in cbind(newdata, as_tibble(indicators)): number of rows of result
#> is not a multiple of vector length (arg 1)

# error step_poly
rec <- recipe(~qux, data = foo) %>% step_poly(qux, options = list(degree = 2))

trained_rec <- prep(rec, training = foo)

# No error: but all
try(bake(trained_rec, newdata = foo))
#> # A tibble: 2 x 0


# error step_bs
rec <- recipe(~qux, data = foo) %>% step_bs(qux)

trained_rec <- prep(rec, training = foo)

try(bake(trained_rec, newdata = foo))
#> Warning in cbind(newdata, as_tibble(bs_values)): number of rows of result
#> is not a multiple of vector length (arg 1)
#> # A tibble: 3 x 0

# error in step_ns
rec <- recipe(~qux, data = foo) %>% step_ns(qux)

trained_rec <- prep(rec, training = foo)

# No error: but all rows dropped
try(bake(trained_rec, newdata = foo))
#> # A tibble: 2 x 0

Session Info:

> sessionInfo()
R version 3.4.3 (2017-11-30)
Platform: x86_64-apple-darwin15.6.0 (64-bit)
Running under: macOS High Sierra 10.13.3

Matrix products: default
BLAS: /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/3.4/Resources/lib/libRlapack.dylib

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] recipes_0.1.2    broom_0.4.3      forcats_0.2.0    stringr_1.2.0   
 [5] dplyr_0.7.4.9000 purrr_0.2.4      readr_1.1.1      tidyr_0.8.0     
 [9] tibble_1.4.2     ggplot2_2.2.1    tidyverse_1.2.1  fortunes_1.5-4  
[13] testthat_2.0.0   rmarkdown_1.8    knitr_1.19       roxygen2_6.0.1  
[17] devtools_1.13.4 

loaded via a namespace (and not attached):
 [1] httr_1.3.1        ddalpha_1.3.1.1   splines_3.4.3     sfsmisc_1.1-1    
 [5] jsonlite_1.5      prodlim_1.6.1     modelr_0.1.1      assertthat_0.2.0 
 [9] DRR_0.0.3         cellranger_1.1.0  yaml_2.1.16       robustbase_0.92-8
[13] ipred_0.9-6       pillar_1.1.0      backports_1.1.2   lattice_0.20-35  
[17] glue_1.2.0        digest_0.6.15     rvest_0.3.2       colorspace_1.3-2 
[21] htmltools_0.3.6   Matrix_1.2-12     plyr_1.8.4        psych_1.7.8      
[25] timeDate_3042.101 pkgconfig_2.0.1   CVST_0.2-1        haven_1.1.1      
[29] scales_0.5.0      gower_0.1.2       lava_1.6          withr_2.1.1      
[33] nnet_7.3-12       lazyeval_0.2.1    cli_1.0.0         mnormt_1.5-5     
[37] survival_2.41-3   magrittr_1.5      crayon_1.3.4      readxl_1.0.0     
[41] memoise_1.1.0     evaluate_0.10.1   nlme_3.1-131      MASS_7.3-48      
[45] xml2_1.2.0        dimRed_0.1.0      foreign_0.8-69    class_7.3-14     
[49] tools_3.4.3       hms_0.4.1         kernlab_0.9-25    munsell_0.4.3    
[53] bindrcpp_0.2      compiler_3.4.3    RcppRoll_0.2.2    rlang_0.1.6.9003 
[57] grid_3.4.3        rstudioapi_0.7    gtable_0.2.0      reshape2_1.4.3   
[61] R6_2.2.2          lubridate_1.7.2   utf8_1.1.3        bindr_0.1        
[65] commonmark_1.4    rprojroot_1.3-2   stringi_1.1.6     parallel_3.4.3   
[69] Rcpp_0.12.15      rpart_4.1-12      DEoptimR_1.0-8    tidyselect_0.2.3 

[topepo]

Thanks for catching this. That does solve the issue. Fire away if you want to submit the PR otherwise I can commit my changes (but if you want the credit...)

[jrnold]

Will do. The changes are already in the linked commits.

[jrnold]

Opened PR #128