discrim (version 0.0.2)

discrim_linear: General Interface for Linear Discriminant Models

Description

discrim_linear() is a way to generate a specification of a linear discriminant analysis (LDA) model before fitting and allows the model to be created using different packages in R.

Usage

discrim_linear(mode = "classification", penalty = NULL)
# S3 method for discrim_linear
update(object, penalty = NULL, fresh = FALSE, ...)

Arguments

mode

A single character string for the type of model. The only possible value for this model is "classification".

penalty

An non-negative number representing the amount of regularization used by some of the engines.

object

A linear discriminant model specification.

fresh

A logical for whether the arguments should be modified in-place of or replaced wholesale.

...

Not used for update().

Engine Details

Engines may have pre-set default arguments when executing the model fit call. For this type of model, the template of the fit calls are:

discrim_linear() %>% 
  set_engine("MASS") %>% 
  translate()

## Linear Discriminant Model Specification (classification)
## 
## Computational engine: MASS 
## 
## Model fit template:
## MASS::lda(formula = missing_arg(), data = missing_arg())

discrim_linear() %>% 
  set_engine("mda") %>% 
  translate()

## Linear Discriminant Model Specification (classification)
## 
## Computational engine: mda 
## 
## Model fit template:
## mda::fda(formula = missing_arg(), data = missing_arg(), method = mda::gen.ridge, 
##     keep.fitted = FALSE)

The standardized parameter names in parsnip can be mapped to their original names in each engine that has main parameters:

parsnip	mda
penalty	lambda

Details

For discrim_linear(), the mode will always be "classification".

The model can be created using the fit() function using the following engines:

• R: "MASS"(the default) or "mda"

The main argument for the model is:

• penalty: The total amount of regularization in the model. Note that this only used for the "mda" engine where it is a pure L2 penalty (a.k.a ridge regression).

This argument is converted to its specific names at the time that the model is fit. Other options and argument can be set using set_engine(). If left to their defaults here (NULL), the values are taken from the underlying model functions. If parameters need to be modified, update() can be used in lieu of recreating the object from scratch.

Examples

# NOT RUN {
parabolic_grid <-
  expand.grid(X1 = seq(-5, 5, length = 100),
              X2 = seq(-5, 5, length = 100))

lda_mod <-
  discrim_linear(penalty = .1) %>%
  set_engine("mda") %>%
  fit(class ~ ., data = parabolic)

parabolic_grid$lda <-
  predict(lda_mod, parabolic_grid, type = "prob")$.pred_Class1

library(ggplot2)
ggplot(parabolic, aes(x = X1, y = X2)) +
  geom_point(aes(col = class), alpha = .5) +
  geom_contour(data = parabolic_grid, aes(z = lda), col = "black", breaks = .5) +
  theme_bw() +
  theme(legend.position = "top") +
  coord_equal()


model <- discrim_linear(penalty = 0.1)
model
update(model, penalty = 1)
# }