plug_model() adds a parsnip model to the tidyflow.
drop_model() removes the model specification as well as any fitted
model object. Any extra formulas are also removed. Doesn't remove any steps
from the pre stage.
replace_model() first removes the model then adds the new specification to
the tidyflow.
plug_model(x, spec, formula = NULL)
drop_model(x)
replace_model(x, spec, formula = NULL)A tidyflow.
A parsnip model specification.
An optional formula override to specify the terms of the model. Typically, the terms are extracted from the formula or recipe preprocessing methods. However, some models (like survival and bayesian models) use the formula not to preprocess, but to specify the structure of the model. In those cases, a formula specifying the model structure must be passed unchanged into the model call itself. This argument is used for those purposes.
x, updated with either a new or removed model.
plug_model() is a required step to construct a minimal tidyflow.
library(parsnip)
# Define two competing model:
lm_model <- set_engine(linear_reg(), "lm")
regularized_model <- set_engine(lm_model, "glmnet")
# Define a minimal tidyflow: data + formula + model
wf <-
mtcars %>%
tidyflow() %>%
plug_formula(mpg ~ .) %>%
plug_model(lm_model)
wf
#> ══ Tidyflow ════════════════════════════════════════════════════════════════════
#> Data: 32 rows x 11 columns
#> Split: None
#> Formula: mpg ~ .
#> Resample: None
#> Grid: None
#> Model:
#> Linear Regression Model Specification (regression)
#>
#> Computational engine: lm
#>
# We can drop the model at any time and the remaining steps
# are intact
drop_model(wf)
#> ══ Tidyflow ════════════════════════════════════════════════════════════════════
#> Data: 32 rows x 11 columns
#> Split: None
#> Formula: mpg ~ .
#> Resample: None
#> Grid: None
#> Model: None
# We can fit the model with `fit`:
fitted <- fit(wf)
# Extract the model if needed:
fitted %>%
pull_tflow_fit()
#> parsnip model object
#>
#>
#> Call:
#> stats::lm(formula = ..y ~ ., data = data)
#>
#> Coefficients:
#> (Intercept) cyl disp hp drat wt
#> 12.30337 -0.11144 0.01334 -0.02148 0.78711 -3.71530
#> qsec vs am gear carb
#> 0.82104 0.31776 2.52023 0.65541 -0.19942
#>
# If we remove the model from the fitted `tidyflow`,
# the fit is dropped:
drop_model(fitted)
#> ══ Tidyflow ════════════════════════════════════════════════════════════════════
#> Data: 32 rows x 11 columns
#> Split: None
#> Formula: mpg ~ .
#> Resample: None
#> Grid: None
#> Model: None
# We could replace the model from the initial tidyflow with
# the regularized model with `replace_model`
## TODO: when https://github.com/cimentadaj/tidyflow/issues/4 is fixed
## replace wf with fitted here.
reg_fitted <-
wf %>%
replace_model(regularized_model) %>%
fit()
#> Error in .check_glmnet_penalty_fit(x): For the glmnet engine, `penalty` must be a single number (or a value of `tune()`).
#> • There are 0 values for `penalty`.
#> • To try multiple values for total regularization, use the tune package.
#> • To predict multiple penalties, use `multi_predict()`
reg_fitted %>%
pull_tflow_fit()
#> Error in is_tidyflow(x): object 'reg_fitted' not found