Package 'RegAssure'

Title: Streamlined Integration of Regression Assumption
Description: It streamlines the evaluation of regression model assumptions, enhancing result reliability. With integrated tools for assessing key aspects like linearity, homoscedasticity, and more. It's a valuable asset for researchers and analysts working with regression models.
Authors: Nicolás Rubio García [aut, cre]
Maintainer: Nicolás Rubio García <[email protected]>
License: GPL (>= 3)
Version: 1.0.0
Built: 2024-10-29 03:46:22 UTC
Source: https://github.com/nrubiog/regassure

Help Index

Check Linear Regression Model Assumptions

Description

This function verifies assumptions for various linear regression models, including OLS, WLS, censored, and trunmessageed models.

Usage

check_lm_assumptions(
  model,
  studentize = TRUE,
  iterations = 1000,
  exact = NULL,
  simulate.p.value = FALSE,
  B = 2000,
  simulate = TRUE
)

Arguments

model

Object of class 'lm'. The fitted linear regression model.
studentize

Argument of bptest function. Default is TRUE.
iterations

Argument of dwtest function. Default is 1000.
exact

Argument of ks.test function. Default is NULL.
simulate.p.value

Argument of ks.test function. Default is FALSE.
B

Argument of ks.test function. Default is 2000.
simulate

Argument of durbinWatsonTest function. Default is TRUE.

Value

A list containing the results of assumption checks.

Check Assumptions for Logit Models

Description

This function performs assumption checks for logistic regression models, including binary, multinomial, and ordered models. It conducts various tests to assess the validity of the model assumptions.

Usage

check_logit(
  logit_model,
  data,
  tipo_modelo,
  vars_numericas = NULL,
  y = NULL,
  auc = NULL,
  ci = NULL,
  ret = NULL
)

Arguments

logit_model

Fitted logistic regression model.
data

dataframe containing predictor variables.
tipo_modelo

Type of logistic regression model: ("binario", "multinomial", "ordenado", "binary", "multinomial", "ordered", "ordinal")
vars_numericas

Numeric variables to be used in Box-Tidwell test. Default is NULL.
y

Response variable for ROC test. Default is NULL.
auc

roc function argument.
ci

roc function argument.
ret

roc function argument.

Value

A list containing the results of assumption checks.

Get Predictions and Errors

Description

This function generates predictions from a model and calculates the errors between actual and predicted values, resulting in a dataframe containing the predicted values, actual values, and the residuals.

Usage

get_predict(modelo, datos, real, n = NULL)

Arguments

modelo

Fitted model object
datos

Data frame containing predictor variables.
real

Vector of actual response values.
n

Number of digits to round the columns. Default

Value

A data frame containing actual, predicted, and error values.