Plot ispline partial effects with quantile and start‐point markers

Given a tidy data frame of ispline basis outputs (from run_ispline_models), this function identifies the ispline column matching a specified covariate, draws thinner ispline curves for each zOrder, and overlays:

• Small symbols at user‐defined quantiles of the raw covariate.

• A larger symbol at each curve’s starting point (minimum covariate value).

Usage

plot_ispline_lines(
  ispline_data,
  x_var,
  orders = NULL,
  cols = NULL,
  shapes = NULL,
  probs = c(0, 0.25, 0.5, 0.75, 1),
  line_size = 0.5,
  point_size = 1.5,
  start_size = 3,
  start_stroke = 0
)

Arguments

ispline_data

A data frame as returned by run_ispline_models, containing raw covariates, their _is spline bases, and a zOrder column.

x_var

A string; the name of the raw covariate to plot (e.g. “dist”). The function will look for a matching spline column named <x_var>_is.

orders

Character vector of zOrder levels (in desired legend/order). Default is unique(ispline_data$zOrder).

cols

Character vector of colours, one per order. Default uses scales::hue_pal().

shapes

Integer vector of plotting symbols (pch codes), one per order. Default is 15:(...).

probs

Numeric vector of probabilities (between 0 and 1) at which to place small quantile markers. Default c(0, .25, .5, .75, 1).

line_size

Numeric; line width for the spline curves. Default 0.5.

point_size

Numeric; size of the quantile markers. Default 1.5.

start_size

Numeric; size of the big start‐point markers. Default 3.

start_stroke

Numeric; stroke width for the big start markers. Default 0.

Value

A ggplot2 object showing:

• geom_line() for each curve.

• geom_point() at the specified quantiles.

• A larger geom_point() at each curve’s minimum.

Details

1. Detects all columns ending in _is and fuzzy‐matches x_var against them.

2. Extracts the raw covariate name by stripping _is.

3. Computes one “start” point per zOrder at the minimum raw covariate.

4. Computes quantile‐closest points per zOrder at the user’s probs.

See also

run_ispline_models, ggplot

Examples

if (FALSE) { # \dontrun{
# load sample data
data(bird.spec.fine)
data(bird.env.fine)

# prepare inputs
xy.bird    <- bird.spec.fine[,1:2]
spp.bird   <- bird.spec.fine[,3:102]
envir.bird <- bird.env.fine[,3:9]

# fit & gather ispline tables
ispline_tabs_all <- run_ispline_models(
  spp_df    = spp.bird,
  env_df    = envir.bird,
  xy_df     = xy.bird,
  orders    = 2:6,
  sam       = 100,
  normalize = "Jaccard",
  reg_type  = "ispline"
)

# Line plot for "dist"
plot_ispline_lines(
  ispline_data = ispline_tabs_all,
  x_var        = "dist",            # will match "dist_is"
  orders       = paste("Order", 2:6),
  cols         = c('green','cyan','purple','blue','black'),
  shapes       = c(15,16,17,18,19)
)
} # }