I have built custom recipes that I use in two of my packages. I suspect the issue may be the same across all. I will start with this one where the issue I am getting is at prep time as it is saying that the variable types are incorrect.
Here is the R file that creates the recipe step function:
step_hai_fourier <- function(recipe,
...,
role = "predictor",
trained = FALSE,
columns = NULL,
scale_type = c("sin", "cos", "sincos"),
period = 1,
order = 1,
skip = FALSE,
id = rand_id("hai_fourier")) {
terms <- recipes::ellipse_check(...)
funcs <- c("sin", "cos", "sincos")
if (!(scale_type %in% funcs)) {
rlang::abort("`func` should be either `sin`, `cos`, or `sincos`")
}
recipes::add_step(
recipe,
step_hai_fourier_new(
terms = terms,
role = role,
trained = trained,
columns = columns,
scale_type = scale_type,
period = period,
order = order,
skip = skip,
id = id
)
)
}
step_hai_fourier_new <-
function(terms, role, trained, columns, scale_type, period, order, skip, id) {
recipes::step(
subclass = "hai_fourier",
terms = terms,
role = role,
trained = trained,
columns = columns,
scale_type = scale_type,
period = period,
order = order,
skip = skip,
id = id
)
}
#' @export
prep.step_hai_fourier <- function(x, training, info = NULL, ...) {
col_names <- recipes::recipes_eval_select(x$terms, training, info)
value_data <- info[info$variable %in% col_names, ]
if (any(value_data$type != "numeric")) {
rlang::abort(
paste0(
"All variables for `step_hai_fourier` must be `numeric`",
"`integer` `double` classes."
)
)
}
step_hai_fourier_new(
terms = x$terms,
role = x$role,
trained = TRUE,
columns = col_names,
scale_type = x$scale_type,
period = x$period,
order = x$order,
skip = x$skip,
id = x$id
)
}
#' @export
bake.step_hai_fourier <- function(object, new_data, ...) {
make_call <- function(col, period, order, scale_type) {
rlang::call2(
"hai_fourier_vec",
.x = rlang::sym(col),
.period = period,
.order = order,
.scale_type = scale_type,
.ns = "healthyR.ai"
)
}
grid <- expand.grid(
col = object$columns,
scale_type = object$scale_type,
period = object$period,
order = object$order,
stringsAsFactors = FALSE
)
calls <- purrr::pmap(.l = list(grid$col, grid$period, grid$order, grid$scale_type), make_call)
# Column Names
newname <- paste0("fourier_", grid$col, "_", grid$scale_type)
calls <- recipes::check_name(calls, new_data, object, newname, TRUE)
tibble::as_tibble(dplyr::mutate(new_data, !!!calls))
}
#' @export
print.step_hai_fourier <-
function(x, width = max(20, options()$width - 35), ...) {
title <- "Fourier Transformation on "
recipes::print_step(
x$columns, x$terms, x$trained,
width = width, title = title
)
invisible(x)
}
#' Requited Packages
#' @rdname required_pkgs.healthyR.ai
#' @keywords internal
#' @return A character vector
#' @param x A recipe step
# @noRd
#' @export
required_pkgs.step_hai_fourier <- function(x, ...) {
c("healthyR.ai")
}
Here is a simple example which is in my man page that is now failing:
len_out <- 10
by_unit <- "month"
start_date <- as.Date("2021-01-01")
data_tbl <- tibble(
date_col = seq.Date(from = start_date, length.out = len_out, by = by_unit),
a = rnorm(len_out),
b = runif(len_out)
)
# Create a recipe object
rec_obj <- recipe(a ~ ., data = data_tbl) %>%
step_hai_fourier(b, scale_type = "sin") %>%
step_hai_fourier(b, scale_type = "cos") %>%
step_hai_fourier(b, scale_type = "sincos")
# View the recipe object
> rec_obj
Recipe
Inputs:
role #variables
outcome 1
predictor 2
Operations:
Fourier Transformation on b
Fourier Transformation on b
Fourier Transformation on b
# Prepare the recipe object
> prep(rec_obj)
Error in `prep()`:
! All variables for `step_hai_fourier` must be `numeric``integer` `double` classes.
Run `rlang::last_error()` to see where the error occurred.
# Bake the recipe object - Adds the Time Series Signature
bake(prep(rec_obj), data_tbl)
rec_obj %>% get_juiced_data()
Here is the trace:
> rlang::last_error()
<error/rlang_error>
Error in `prep()`:
! All variables for `step_hai_fourier` must be `numeric``integer` `double` classes.
---
Backtrace:
1. recipes::prep(rec_obj)
2. recipes:::prep.recipe(rec_obj)
4. healthyR.ai:::prep.step_hai_fourier(x$steps[[i]], training = training, info = x$term_info)
Run `rlang::last_trace()` to see the full context.
> rlang::last_trace()
<error/rlang_error>
Error in `prep()`:
! All variables for `step_hai_fourier` must be `numeric``integer` `double` classes.
---
Backtrace:
▆
1. ├─recipes::prep(rec_obj)
2. └─recipes:::prep.recipe(rec_obj)
3. ├─recipes::prep(x$steps[[i]], training = training, info = x$term_info)
4. └─healthyR.ai:::prep.step_hai_fourier(x$steps[[i]], training = training, info = x$term_info)
5. └─rlang::abort(...)
Package requires recipes >= 1.0.0 and I have no clue what happened.
SESSION INFO
> sessionInfo()
R version 4.2.1 (2022-06-23 ucrt)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 19044)
Matrix products: default
locale:
[1] LC_COLLATE=English_United States.utf8 LC_CTYPE=English_United States.utf8
[3] LC_MONETARY=English_United States.utf8 LC_NUMERIC=C
[5] LC_TIME=English_United States.utf8
attached base packages:
[1] stats graphics grDevices utils datasets methods base
other attached packages:
[1] tidyquant_1.0.5 quantmod_0.4.20
[3] TTR_0.24.3 PerformanceAnalytics_2.0.4
[5] xts_0.12.2 zoo_1.8-11
[7] lubridate_1.9.0 timechange_0.1.1
[9] timetk_2.8.1 healthyR.ai_0.0.9
[11] yardstick_1.1.0 workflowsets_1.0.0
[13] workflows_1.1.0 tune_1.0.1
[15] rsample_1.1.0 recipes_1.0.3
[17] parsnip_1.0.3 modeldata_1.0.1
[19] infer_1.0.3 dials_1.1.0
[21] scales_1.2.1 broom_1.0.1
[23] tidymodels_1.0.0 forcats_0.5.2
[25] stringr_1.4.1 dplyr_1.0.10
[27] purrr_0.3.5 readr_2.1.3
[29] tidyr_1.2.1 tibble_3.1.8
[31] ggplot2_3.4.0 tidyverse_1.3.2
loaded via a namespace (and not attached):
[1] googledrive_2.0.0 colorspace_2.0-3 ellipsis_0.3.2 class_7.3-20
[5] fs_1.5.2 rstudioapi_0.14 listenv_0.8.0 furrr_0.3.1
[9] prodlim_2019.11.13 fansi_1.0.3 xml2_1.3.3 codetools_0.2-18
[13] splines_4.2.1 knitr_1.40 jsonlite_1.8.3 dbplyr_2.2.1
[17] compiler_4.2.1 httr_1.4.4 backports_1.4.1 assertthat_0.2.1
[21] Matrix_1.5-3 gargle_1.2.1 cli_3.4.1 tools_4.2.1
[25] gtable_0.3.1 glue_1.6.2 Rcpp_1.0.9 fracdiff_1.5-2
[29] cellranger_1.1.0 DiceDesign_1.9 vctrs_0.5.0 nlme_3.1-160
[33] urca_1.3-3 iterators_1.0.14 lmtest_0.9-40 timeDate_4021.106
[37] gower_1.0.0 xfun_0.34 globals_0.16.1 rvest_1.0.3
[41] lifecycle_1.0.3 googlesheets4_1.0.1 future_1.29.0 MASS_7.3-58.1
[45] ipred_0.9-13 hms_1.1.2 parallel_4.2.1 curl_4.3.3
[49] padr_0.6.1 rpart_4.1.19 stringi_1.7.8 highr_0.9
[53] tseries_0.10-52 foreach_1.5.2 lhs_1.1.5 hardhat_1.2.0
[57] lava_1.7.0 rlang_1.0.6 pkgconfig_2.0.3 evaluate_0.18
[61] lattice_0.20-45 tidyselect_1.2.0 parallelly_1.32.1 magrittr_2.0.3
[65] R6_2.5.1 generics_0.1.3 DBI_1.1.3 pillar_1.8.1
[69] haven_2.5.1 withr_2.5.0 survival_3.4-0 nnet_7.3-18
[73] future.apply_1.10.0 modelr_0.1.10 crayon_1.5.2 Quandl_2.11.0
[77] utf8_1.2.2 tzdb_0.3.0 grid_4.2.1 readxl_1.4.1
[81] forecast_8.18 reprex_2.0.2 digest_0.6.30 GPfit_1.0-8
[85] munsell_0.5.0 sessioninfo_1.2.2 quadprog_1.5-8
RStudio Version
RStudio 2022.07.2+576 "Spotted Wakerobin" Release (e7373ef832b49b2a9b88162cfe7eac5f22c40b34, 2022-09-06) for Windows
Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) QtWebEngine/5.12.8 Chrome/69.0.3497.128 Safari/537.36
