Skip to content

Fit Multivariate Logistic Regression Model

fit_logistic.Rd

A wrapper around glm() for fitting multivariate logistic regression models for binary classification problems.

Usage

fit_logistic(x, ..., strip)

# S3 method for class 'formula'
fit_logistic(formula, ..., strip = FALSE)

# S3 method for class 'data.frame'
fit_logistic(x, y = NULL, strip = FALSE, ...)

# S3 method for class 'tr_data'
fit_logistic(x, ..., strip = FALSE)

# S3 method for class 'matrix'
fit_logistic(x, y, strip = FALSE, ...)

Arguments

strip

logical(1). Should certain entries of the model object be stripped via stripLMC() to reduce object size? If true, some downstream functionality is compromised, e.g. summary() and residuals(), however when iterating over 1000s of models this may be an acceptable trade-off to limit runaway memory consumption.

formula, x, ...

Either a formula, data frame, or matrix. If a formula (preferred) should be a model of the form: \(class ~ x_1 + x_2 + ... + x_n\). If a data frame (preferably a tr_data object), containing features or predictors. If a matrix object containing ONLY predictors, in which case y must be passed (see examples below). Unmatched arguments eventually be passed to glm() via the ....

y

Can be one of two options:

character

A character(1) indicating the column in x containing the true class names.

vector

A vector factor(n) of true class names for each sample (row) in x.

Value

A glm model object as returned by glm(), logistic regression model.

Methods (by class)

  • fit_logistic(formula): S3 formula method for fit.

  • fit_logistic(data.frame): S3 data.frame method for fit_logistic.

  • fit_logistic(tr_data): S3 tr_data method for fit_logistic.

  • fit_logistic(matrix): S3 matrix method for fit_logistic.

See also

glm()

The fit*() family: fit_gbm(), fit_kknn(), fit_nb()

Author

Stu Field

Examples

# formula S3 method
# This is the preferred syntax
class(tr_iris)
#> [1] "tr_data"    "tbl_df"     "tbl"        "data.frame"

df <- tibble::as_tibble(tr_iris)  # strip tr_data class

# tr_data S3 method:
model <- fit_logistic(tr_iris)

# data frame S3 method:
model <- fit_logistic(df, "Species")

# formula S3 method:
model <- fit_logistic(Species ~ ., data = df)

# data frame S3 method (2):
model <- fit_logistic(df[, -5L], y = df$Species)  # vector of class names

# matrix S3 method:
model <- fit_logistic(as.matrix(df[, -5L]), y = df$Species)  # 'glmnet' syntax