Scatter Plot with Marginal Density/Histogram and Regression Equation (Enhanced Version)

Main plot: scatter points + linear regression line + automatic annotation of equation, R^2, p value (ggpmisc::stat_poly_eq). Marginal plots: bilateral densigram (density + histogram mix), supports grouping colors. Color logic: uniform color when no color grouping; automatic coloring by group variable with hidden legend when color specified.

Usage

scatter_lm_marginal(
  data,
  x,
  y,
  digits = 3,
  label.x = NULL,
  label.y = NULL,
  color = NULL,
  marginal = c("histogram", "density", "boxplot"),
  bins = 15,
  margin_fill = "grey60",
  margin_color = "white",
  ...
)

Arguments

data: Data frame, must contain x, y and optional color grouping variable
x, y: Character scalars, variable names
digits: Decimal places for equation/statistics, default 3 (controlled by stat_poly_eq)
label.x, label.y: Deprecated, kept for backward compatibility
color: Character scalar, grouping variable name; NULL for no grouping
marginal: Marginal plot type, fixed as "densigram"
bins: Number of histogram bins, passed to ggExtra::ggMarginal
margin_fill, margin_color: Fill color and border color for marginal plots
...: Additional aesthetic parameters passed to geom_point()

Value

    ggplot object (with marginal layers), can be further customized with + or saved

Details

Uses ggpmisc::stat_poly_line() + stat_poly_eq() to automatically add regression line and equation
Marginal plots are densigram (density + histogram), supports group colors/fill
Uniform color #479E88 when no grouping; uses scale_color_npg() with hidden legend when grouping
Returned object already includes marginal layers, can be directly saved with ggsave()

Examples

if (FALSE) { # \dontrun{
p <- scatter_lm_marginal(mtcars, "wt", "mpg", color = "cyl", bins = 20)
print(p)
ggsave("scatter_marginal.png", p, width = 6, height = 5)
} # }