Linear regression fitting and formatting helpers

Usage

leo_linear(df, y_out, x_exp, x_cov = NULL, x_exp_type = "auto", verbose = TRUE)

leo_linear_format(x, style = "wide")

Arguments

df: Data frame containing the continuous outcome, exposure, and covariates.
y_out: Character scalar giving the continuous outcome column name.
x_exp: Character scalar giving the exposure column name.
x_cov: NULL, a character vector of covariate column names, or a list of covariate column-name vectors. All models are fitted on the same complete-case cohort defined by y_out, x_exp, and all covariates that appear in x_cov. If x_cov is a named list, those model names are preserved in the output columns.
x_exp_type: Exposure type handling for x_exp. Use "auto" to infer from the input type, "continuous" to force a numeric linear term, or "categorical" to force factor coding.
verbose: Logical; print progress messages.
x: Result returned by leo_linear().
style: One of "wide", "tidy", or "gtsummary".

Value

A leo_linear object containing the default wide table in $result, the underlying tidy rows in $result_tidy, model metadata, and fitted lm objects.

Details

leo_linear() fits one or more linear regression models for a single exposure and continuous outcome. leo_linear_format() converts the returned object into a wide summary table, tidy result table, or gtsummary output.

Examples

linear_df <- dplyr::transmute(
  mtcars,
  outcome = mpg,
  wt = wt,
  am = factor(am, levels = c(0, 1), labels = c("Auto", "Manual")),
  cyl_group = factor(cyl, levels = c(4, 6, 8), labels = c("Cyl4", "Cyl6", "Cyl8"))
)

model_cont <- list(
  "Crude" = NULL,
  "Model A" = c("am"),
  "Model B" = c("am", "cyl_group")
)

res_wt <- leo_linear(
  df = linear_df, y_out = "outcome", x_exp = "wt",
  x_cov = model_cont, verbose = FALSE
)
res_wt$result
#>   Exposure Outcome  N      Class Crude Beta   Crude 95% CI Crude P value
#> 1       wt outcome 32 Continuous     -5.344 -6.486, -4.203        <0.001
#>   Model A Beta Model A 95% CI Model A P value Model B Beta Model B 95% CI
#> 1       -5.353 -6.965, -3.741          <0.001        -3.15 -5.013, -1.286
#>   Model B P value
#> 1           0.002
leo_linear_format(res_wt, style = "tidy")
#>     Model Exposure Outcome Level  N   Beta         95% CI P value      Class
#> 1   Crude       wt outcome  <NA> 32 -5.344 -6.486, -4.203  <0.001 Continuous
#> 2 Model A       wt outcome  <NA> 32 -5.353 -6.965, -3.741  <0.001 Continuous
#> 3 Model B       wt outcome  <NA> 32 -3.150 -5.013, -1.286   0.002 Continuous
#>                               Formula
#> 1                  outcome ~ exposure
#> 2             outcome ~ exposure + am
#> 3 outcome ~ exposure + am + cyl_group

model_cat <- list(
  "Crude" = NULL,
  "Model A" = c("am")
)

res_cyl <- leo_linear(
  df = linear_df, y_out = "outcome", x_exp = "cyl_group",
  x_cov = model_cat, x_exp_type = "categorical", verbose = FALSE
)
res_cyl$result
#>     Exposure Outcome  N                  Class Crude Beta    Crude 95% CI
#> 1 Cyl4 (Ref) outcome 11 Categorical (3 levels)      0.000    0.000, 0.000
#> 2       Cyl6 outcome  7 Categorical (3 levels)     -6.921 -10.108, -3.734
#> 3       Cyl8 outcome 14 Categorical (3 levels)    -11.564 -14.220, -8.908
#>   Crude P value Model A Beta  Model A 95% CI Model A P value
#> 1            NA        0.000    0.000, 0.000              NA
#> 2        <0.001       -6.156  -9.302, -3.010          <0.001
#> 3        <0.001      -10.068 -13.042, -7.093          <0.001

if (requireNamespace("gtsummary", quietly = TRUE) && requireNamespace("broom.helpers", quietly = TRUE)) {
  leo_linear_format(res_wt, style = "gtsummary")
}