Skip to contents

Plot Landscape Metrics

Source: R/plot_metrics.R
plot_metrics.Rd

Creates a visualization of landscape metric values across landscape types using boxplots with overlaid jittered points. Metrics are displayed in separate facets.

Usage

plot_metrics(
  metrics,
  selected_metrics = NULL,
  force = FALSE,
  metric_labels = "abbreviation",
  label_wrap_width = NULL,
  pattern_order = NULL,
  jitter_seed = NA,
  jitter_width = 0.1,
  point_size = 1,
  point_alpha = 0.7
)

Arguments

metrics

Data frame from calculate_metrics. Must contain columns: "level", "pattern", "metric", and "value". For class-level metrics, must also contain "class".

selected_metrics

Character vector of metric abbreviations to visualize, as they appear in the metric column, or the object returned by evaluate_metrics. Must be present in the metrics data. If NULL (default), all available metrics are plotted in alphabetical order, subject to automatic limits based on the number of patterns. Use list_lsm to look up what an abbreviation stands for, or set metric_labels = "name" to show the full names.

force

Logical. Override automatic metric limits (default: FALSE). When TRUE, all selected metrics will be plotted regardless of readability.

metric_labels

Character string controlling how metrics are labelled in facet strips. One of "abbreviation" (default) to use the metric abbreviations as they appear in metrics (e.g. "ai"), or "name" to use the full metric names from list_lsm (e.g. "Aggregation index"). Metrics that summarise per-patch values get the statistic in brackets, since list_lsm() gives the _cv/_mn/_sd triple a single name: area_mn becomes "Patch area (mean)". For class-level metrics the class is added too (e.g. "Patch area (mean, class 1)"), and the metrics data must also contain the metric_name column produced by calculate_metrics.

label_wrap_width

Integer or NULL (default). Character width at which to wrap full metric names in facet strips. Only used when metric_labels = "name". If NULL, a width is chosen automatically from the number of facet columns. The choice wraps by character count rather than rendered text width, so set it explicitly if your font, figure size, or metric selection needs something different.

pattern_order

Character vector giving the order in which patterns should appear along the y-axis, or NULL (default) for alphabetical order. Must contain exactly the patterns present in metrics, i.e. every pattern once and no others. The first element is drawn at the bottom of the axis.

jitter_seed

Seed controlling the random jitter of the points, passed to position_jitter. NA (default) draws fresh jitter each render.

jitter_width

Numeric. Horizontal spread of the points around each pattern (default: 0.1). Set to 0 to disable jitter. Points are never displaced along the value axis, so they always sit at their true metric value.

point_size

Numeric. Size of the data points (default: 1). Reduce for plots with many landscapes per pattern.

point_alpha

Numeric between 0 and 1. Opacity of the data points (default: 0.7). Reduce to make overlapping points easier to read.

Value

A ggplot2 object showing boxplots of metric values by pattern type.

Details

The function automatically limits the number of metrics based on the number of patterns to maintain readability: - 1-3 patterns: up to 12 metrics (3 rows x 4 columns) - 4-5 patterns: up to 8 metrics (2 rows x 4 columns) - 6+ patterns: up to 6 metrics (2 rows x 3 columns)

See also

calculate_metrics, evaluate_metrics, list_lsm for the available metrics and their full names

Other visualization: plot_classified_landscapes(), plot_landscapes()

Examples

landscapes <- create_landscapes(n = 8, patterns = c("labyrinth", "spots"))
#> ✔ Successfully generated all 8 training landscapes
metrics <- calculate_metrics(landscapes, level = "landscape")
#>  ■■■■■■■■■■■■■■■■                  52% |  ETA:  3s
plot_metrics(metrics, selected_metrics = c("ai", "lsi"))


# With more metrics than fit the grid, automatic limiting applies
many_metrics <- c("ai", "lsi", "ed", "np", "pd", "cohesion", "division",
                  "split", "mesh", "enn_mn", "area_mn", "core_mn",
                  "para_mn")
plot_metrics(metrics, selected_metrics = many_metrics)
#> Warning: With 2 patterns, limiting to 12 of 13 requested metrics for readability.
#> ℹ Showing: "ai", "lsi", "ed", "np", "pd", "cohesion", "division", "split",
#>   "mesh", "enn_mn", "area_mn", and "core_mn"
#> ℹ Use `force = TRUE` to show all metrics.
#> Warning: Removed 2 rows containing non-finite outside the scale range
#> (`stat_boxplot()`).
#> Warning: Removed 2 rows containing missing values or values outside the scale range
#> (`geom_point()`).


# Override limits if needed
plot_metrics(metrics, selected_metrics = many_metrics, force = TRUE)
#> Warning: Removed 2 rows containing non-finite outside the scale range
#> (`stat_boxplot()`).
#> Warning: Removed 2 rows containing missing values or values outside the scale range
#> (`geom_point()`).


# Use full metric names instead of abbreviations in facet labels
plot_metrics(metrics, selected_metrics = c("ai", "lsi"), metric_labels = "name")


# Override the automatic wrap width for full metric names
plot_metrics(
  metrics,
  selected_metrics = c("ai", "lsi"),
  metric_labels = "name",
  label_wrap_width = 15
)


# Control the order of patterns on the y-axis
plot_metrics(
  metrics,
  selected_metrics = c("ai", "lsi"),
  pattern_order = c("spots", "labyrinth")
)


# Fix the jitter so that repeated runs produce an identical figure
plot_metrics(metrics, selected_metrics = c("ai", "lsi"), jitter_seed = 42)


# Adjust point appearance for plots with many landscapes per pattern
plot_metrics(
  metrics,
  selected_metrics = c("ai", "lsi"),
  point_size = 0.5,
  point_alpha = 0.4
)