Alice Hobbs
Reputation: 1215
Tidymodels: Decision Tree Learning in R - Error: No variables or terms were selected
Overview:
I have a data frame called 'FID'and I am attempting to follow this tutorial (see below) to produce three models: (1) Bagged trees; (2) Random Forests; and (3) Boosted Trees.
Tutorial:
https://bcullen.rbind.io/post/2020-06-02-tidymodels-decision-tree-learning-in-r/
Issue
when I try to run the models, "fit_bag", fit_rf, "fit_boost", I am experiencing this error message below. I think the issue may lie in the pre-processing phase.
Would anyone be able to help advise with this issue?
Many thanks in advance.
Error - no variables are selected
i Fold01: recipe
x Fold01: recipe: Error: No variables or terms were selected.
i Fold02: recipe
x Fold02: recipe: Error: No variables or terms were selected.
i Fold03: recipe
x Fold03: recipe: Error: No variables or terms were selected.
i Fold04: recipe
x Fold04: recipe: Error: No variables or terms were selected.
i Fold05: recipe
x Fold05: recipe: Error: No variables or terms were selected.
i Fold06: recipe
x Fold06: recipe: Error: No variables or terms were selected.
i Fold07: recipe
x Fold07: recipe: Error: No variables or terms were selected.
i Fold08: recipe
x Fold08: recipe: Error: No variables or terms were selected.
i Fold09: recipe
x Fold09: recipe: Error: No variables or terms were selected.
i Fold10: recipe
x Fold10: recipe: Error: No variables or terms were selected.
Warning message:
All models failed in [fit_resamples()]. See the `.notes` column.
R-code
##Open library packages
library(tidymodels)
library(tidyverse) # manipulating data
library(skimr) # data visualization
library(baguette) # bagged trees
library(future) # parallel processing & decrease computation time
library(xgboost) # boosted trees
# split the data
split <- initial_split(Tidmodel_df)
# extract the training data
train <- training(split)
# resample the data with 10-fold cross-validation (10-fold by default)
cv <- vfold_cv(train)
##Preprocessing
rec <- recipe(Frequency~., data=train) %>%
update_role(contains("id"), Year, Month, Monsoon, Days, new_role = "id vars") %>% # declare ID variables
step_nzv(all_predictors(), freq_cut = 0, unique_cut = 0) %>% # remove variables with zero variances
step_novel(all_nominal()) %>% # prepares test data to handle previously unseen factor levels
step_medianimpute(all_numeric(), -all_outcomes(), -has_role("id vars")) %>% # replaces missing numeric observations with the median
step_dummy(all_nominal(), -has_role("id vars")) # dummy codes categorical variables
###########################################################
##Create Models
###########################################################
#####Bagged Trees
mod_bag <- bag_tree() %>%
set_mode("regression") %>%
set_engine("rpart", times = 10) # 10 bootstrap resamples
##Create workflow
wflow_bag <- workflow() %>%
add_recipe(rec) %>%
add_model(mod_bag)
##Fit the model
plan(multisession)
fit_bag <- fit_resamples(
wflow_bag,
cv,
metrics = metric_set(rmse, rsq),
control = control_resamples(verbose = TRUE,
save_pred = TRUE,
extract = function(x) extract_model(x)))
##Random forests
mod_rf <-rand_forest() %>%
set_engine("ranger",
num.threads = parallel::detectCores(),
importance = "permutation",
verbose = TRUE) %>%
set_mode("regression") %>%
set_args(trees = 1000)
##Create Workflow
wflow_rf <- workflow() %>%
add_model(mod_rf) %>%
add_recipe(rec)
##Fit the model
plan(multisession)
fit_rf <- fit_resamples(
wflow_rf,
cv,
metrics = metric_set(rmse, rsq),
control = control_resamples(verbose = TRUE,
save_pred = TRUE,
extract = function(x) x)
)
##Boosted Trees
mod_boost <- boost_tree() %>%
set_engine("xgboost", nthreads = parallel::detectCores()) %>%
set_mode("regression")
##Create workflow
wflow_boost <- workflow() %>%
add_recipe(rec) %>%
add_model(mod_boost)
##Fit model
plan(multisession)
fit_boost <- fit_resamples(
wflow_boost,
cv,
metrics = metric_set(rmse, rsq),
control = control_resamples(verbose = TRUE,
save_pred = TRUE)
)
Data frame - FID
structure(list(Year = c(2015, 2015, 2015, 2015, 2015, 2015, 2015,
2015, 2015, 2015, 2015, 2015, 2016, 2016, 2016, 2016, 2016, 2016,
2016, 2016, 2016, 2016, 2016, 2016, 2017, 2017, 2017, 2017, 2017,
2017, 2017, 2017, 2017, 2017, 2017, 2017), Month = structure(c(1L,
2L, 3L, 4L, 5L, 6L, 7L, 8L, 9L, 10L, 11L, 12L, 1L, 2L, 3L, 4L,
5L, 6L, 7L, 8L, 9L, 10L, 11L, 12L, 1L, 2L, 3L, 4L, 5L, 6L, 7L,
8L, 9L, 10L, 11L, 12L), .Label = c("January", "February", "March",
"April", "May", "June", "July", "August", "September", "October",
"November", "December"), class = "factor"), Monsoon = structure(c(2L,
2L, 1L, 1L, 4L, 4L, 4L, 4L, 4L, 3L, 3L, 2L, 2L, 2L, 1L, 1L, 4L,
4L, 4L, 4L, 4L, 3L, 3L, 2L, 2L, 2L, 1L, 1L, 4L, 4L, 4L, 4L, 4L,
3L, 3L, 2L), .Label = c("First_Inter_Monssoon", "North_Monsoon",
"Second_Inter_Monsoon", "South_Monsson"), class = "factor"),
Frequency = c(36, 28, 39, 46, 5, 0, 0, 22, 10, 15, 8,
33, 33, 29, 31, 23, 8, 9, 7, 40, 41, 41, 30, 30, 44, 37,
41, 42, 20, 0, 7, 27, 35, 27, 43, 38), Days = c(31,
28, 31, 30, 6, 0, 0, 29, 15, 29, 29, 31, 31, 29, 30, 30,
7, 0, 7, 30, 30, 31, 30, 27, 31, 28, 30, 30, 21, 0, 7, 26,
29, 27, 29, 29)), row.names = c(NA, -36L), class = "data.frame")
Upvotes: 1
Views: 320
Answers (1)
Julia Silge
Reputation: 11623
The problem here is when you use update_role(contains("id"), Year, Month, Monsoon, Days, new_role = "id vars"), you update the role of all the variables such as Year, Month, Monsoon, etc to be "id vars" and then they aren't predictors anymore. When the recipe moves to the next preprocessing steps, it finds that there aren't any predictors at all.
If you want to use those variables as predictors, then leave their roles as is and do not change them to something else like "id vars":
library(tidymodels)
library(baguette) # bagged trees
fid_df <- structure(list(Year = c(2015, 2015, 2015, 2015, 2015, 2015, 2015,
2015, 2015, 2015, 2015, 2015, 2016, 2016, 2016, 2016, 2016, 2016,
2016, 2016, 2016, 2016, 2016, 2016, 2017, 2017, 2017, 2017, 2017,
2017, 2017, 2017, 2017, 2017, 2017, 2017), Month = structure(c(1L,
2L, 3L, 4L, 5L, 6L, 7L, 8L, 9L, 10L, 11L, 12L, 1L, 2L, 3L, 4L,
5L, 6L, 7L, 8L, 9L, 10L, 11L, 12L, 1L, 2L, 3L, 4L, 5L, 6L, 7L,
8L, 9L, 10L, 11L, 12L), .Label = c("January", "February", "March",
"April", "May", "June", "July", "August", "September", "October",
"November", "December"), class = "factor"), Monsoon = structure(c(2L,
2L, 1L, 1L, 4L, 4L, 4L, 4L, 4L, 3L, 3L, 2L, 2L, 2L, 1L, 1L, 4L,
4L, 4L, 4L, 4L, 3L, 3L, 2L, 2L, 2L, 1L, 1L, 4L, 4L, 4L, 4L, 4L,
3L, 3L, 2L), .Label = c("First_Inter_Monssoon", "North_Monsoon",
"Second_Inter_Monsoon", "South_Monsson"), class = "factor"),
Frequency = c(36, 28, 39, 46, 5, 0, 0, 22, 10, 15, 8,
33, 33, 29, 31, 23, 8, 9, 7, 40, 41, 41, 30, 30, 44, 37,
41, 42, 20, 0, 7, 27, 35, 27, 43, 38), Days = c(31,
28, 31, 30, 6, 0, 0, 29, 15, 29, 29, 31, 31, 29, 30, 30,
7, 0, 7, 30, 30, 31, 30, 27, 31, 28, 30, 30, 21, 0, 7, 26,
29, 27, 29, 29)), row.names = c(NA, -36L), class = "data.frame")
# split the data
fid_split <- initial_split(fid_df)
# extract the training data
fid_train <- training(fid_split)
# resample the data with 10-fold cross-validation (10-fold by default)
cv <- vfold_cv(fid_train)
##Preprocessing
rec <- recipe(Frequency ~ ., data = fid_df) %>%
step_nzv(all_predictors(), freq_cut = 0, unique_cut = 0) %>% # remove variables with zero variances
step_novel(all_nominal()) %>% # prepares test data to handle previously unseen factor levels
step_medianimpute(all_numeric(), -all_outcomes(), -has_role("id vars")) %>% # replaces missing numeric observations with the median
step_dummy(all_nominal(), -has_role("id vars")) # dummy codes categorical variables
rf_spec <- rand_forest(trees = 1e3) %>%
set_engine("ranger", importance = "permutation") %>%
set_mode("regression")
wflow_rf <- workflow() %>%
add_model(rf_spec) %>%
add_recipe(rec)
fit_resamples(
wflow_rf,
cv,
metrics = metric_set(rmse, rsq),
control = control_resamples(save_pred = TRUE)
)
#>
#> Attaching package: 'rlang'
#> The following objects are masked from 'package:purrr':
#>
#> %@%, as_function, flatten, flatten_chr, flatten_dbl, flatten_int,
#> flatten_lgl, flatten_raw, invoke, list_along, modify, prepend,
#> splice
#>
#> Attaching package: 'vctrs'
#> The following object is masked from 'package:tibble':
#>
#> data_frame
#> The following object is masked from 'package:dplyr':
#>
#> data_frame
#> # Resampling results
#> # 10-fold cross-validation
#> # A tibble: 10 x 5
#> splits id .metrics .notes .predictions
#> <list> <chr> <list> <list> <list>
#> 1 <split [24/3]> Fold01 <tibble [2 × 4]> <tibble [0 × 1]> <tibble [3 × 4]>
#> 2 <split [24/3]> Fold02 <tibble [2 × 4]> <tibble [0 × 1]> <tibble [3 × 4]>
#> 3 <split [24/3]> Fold03 <tibble [2 × 4]> <tibble [0 × 1]> <tibble [3 × 4]>
#> 4 <split [24/3]> Fold04 <tibble [2 × 4]> <tibble [0 × 1]> <tibble [3 × 4]>
#> 5 <split [24/3]> Fold05 <tibble [2 × 4]> <tibble [0 × 1]> <tibble [3 × 4]>
#> 6 <split [24/3]> Fold06 <tibble [2 × 4]> <tibble [0 × 1]> <tibble [3 × 4]>
#> 7 <split [24/3]> Fold07 <tibble [2 × 4]> <tibble [0 × 1]> <tibble [3 × 4]>
#> 8 <split [25/2]> Fold08 <tibble [2 × 4]> <tibble [0 × 1]> <tibble [2 × 4]>
#> 9 <split [25/2]> Fold09 <tibble [2 × 4]> <tibble [0 × 1]> <tibble [2 × 4]>
#> 10 <split [25/2]> Fold10 <tibble [2 × 4]> <tibble [0 × 1]> <tibble [2 × 4]>
Created on 2020-11-18 by the reprex package (v0.3.0.9001)
Upvotes: 1
Related Questions
- Error on running `predict()` in tidymodels "Error in `dplyr::select()`: Can't subset columns that don't exist."
- How to make predictions in tidymodels R when feature selection has been applied to the model
- Tidymodels, All models failed; error in model.frame.default and prediction from a rank-deficient fit may be misleading
- Error: "Missing data in columns" when using tidymodels workflow to predict test set
- tidymodel error, when calling predict function is asking for target variable
- Error: Can't subset columns that don't exist when running prediction using {Tidymodels}
- Tidymodels (Fitting Bagged Trees with 10-Fold Cross Validation in R): x Fold01: model: Error: Input must be a vector, not NULL
- Tidymodel Package: General linear models (glm) and decision tree (bagged trees, boosted trees, and random forest) models in R
- R: decision tree Error in `[.data.frame`(frame, predictors) : undefined columns selected
- prediction_regression_tree_ caret: Error in '[.data.frame'(m, labs) : undefined columns selected