| Title: | Odds Plot to Visualise a Logistic Regression Model |
|---|---|
| Description: | Uses the outputs of a logistic regression model, from caret <https://CRAN.R-project.org/package=caret>, to build an odds plot. This allows for the rapid visualisation of odds plot ratios and works best with the outputs of CARET's GLM model class, by returning the final trained model. |
| Authors: | Gary Hutson [aut, cre] |
| Maintainer: | Gary Hutson <[email protected]> |
| License: | GPL (>= 3) |
| Version: | 1.0.2 |
| Built: | 2025-01-23 04:40:53 UTC |
| Source: | https://github.com/statsgary/oddsplotty |
This has been created to generate odds plots on the back of results from a generalised linear model.
odds_plot( x, x_label = "Variables", y_label = "Odds Ratio", title = NULL, subtitle = NULL, point_col = "blue", error_bar_colour = "black", point_size = 5, error_bar_width = 0.3, h_line_color = "black" )odds_plot( x, x_label = "Variables", y_label = "Odds Ratio", title = NULL, subtitle = NULL, point_col = "blue", error_bar_colour = "black", point_size = 5, error_bar_width = 0.3, h_line_color = "black" )
x |
The trained caret GLM logistic regression model |
x_label |
The label name for the x_label |
y_label |
The label name for the y_label |
title |
Title for the Odds Plot |
subtitle |
Subtitle for the Odds Plot |
point_col |
Defaults to blues, but R colour codes can be passed |
error_bar_colour |
the colour of the error bar |
point_size |
the point size of the plot |
error_bar_width |
the width of the displayed error bar |
h_line_color |
the colour of the horizontal line |
A list of the odds returned from logistic regression and a plot showing the odds
#We will use the cancer dataset to build a GLM model to predict cancer status #this will detail whether the patient has a benign or malignant library(mlbench) library(caret) library(tibble) library(ggplot2) library(OddsPlotty) library(e1071) library(ggthemes) #Bring in the data data("BreastCancer", package = "mlbench") breast <- BreastCancer[complete.cases(BreastCancer), ] breast <- breast[, -1] head(breast, 10) breast$Class <- factor(breast$Class) for(i in 1:9) { breast[, i] <- as.numeric(as.character(breast[, i])) } #Train GLM model glm_model <- train(Class ~ ., data = breast, method = "glm", family = "binomial") #Visualise the data with OddsPlotty plotty <- OddsPlotty::odds_plot(glm_model$finalModel,title = "Odds Plot") plotty$odds_plot #Extract underlying odds ratios plotty$odds_data#We will use the cancer dataset to build a GLM model to predict cancer status #this will detail whether the patient has a benign or malignant library(mlbench) library(caret) library(tibble) library(ggplot2) library(OddsPlotty) library(e1071) library(ggthemes) #Bring in the data data("BreastCancer", package = "mlbench") breast <- BreastCancer[complete.cases(BreastCancer), ] breast <- breast[, -1] head(breast, 10) breast$Class <- factor(breast$Class) for(i in 1:9) { breast[, i] <- as.numeric(as.character(breast[, i])) } #Train GLM model glm_model <- train(Class ~ ., data = breast, method = "glm", family = "binomial") #Visualise the data with OddsPlotty plotty <- OddsPlotty::odds_plot(glm_model$finalModel,title = "Odds Plot") plotty$odds_plot #Extract underlying odds ratios plotty$odds_data