Skip to contents

Calculate Landscape Metrics

Source: R/metrics.R
calculate_metrics.Rd

Calculates selected landscape metrics for one or more landscapes using functions from the landscapemetrics package. Returns a standardized tibble with results including landscape identifiers, metric values, and any warnings.

Usage

calculate_metrics(landscapes, metrics = NULL, level = "landscape")

Arguments

landscapes

A single landscape object (created with create_landscape) or a list of landscape objects (e.g. created with create_landscapes). Each landscape object should contain a data element with a SpatRaster, plus name and pattern metadata.

metrics

Character vector. Abbreviations of the metrics to calculate (default: NULL for all available metrics at the specified level). Use list_lsm to look up the available abbreviations and the full metric name each one stands for.

level

Character. Metric level to calculate: "class" or "landscape" (default).

Value

A tibble with the following columns:

landscape_id

Numeric identifier for each landscape in the input list

landscape_name

Name of the landscape from the landscape object

pattern

Pattern type from the landscape object (e.g., "labyrinth", "spots")

layer

Layer number (from landscapemetrics output)

level

Metric level: "class", or "landscape"

class

Class value (for class-level metrics, NA for landscape-level)

metric_name

Metric abbreviation without the class suffix, e.g. "ai". Identical to metric for landscape-level metrics.

metric

Metric abbreviation identifying the row, e.g. "ai". For class-level metrics the class is appended, e.g. "ai_1", so that each class gets its own identifier.

value

Calculated metric value

warnings

Any warnings generated during calculation (NA if none)

n_row, n_col

Cell dimensions of the landscape the row was computed from

cell_size_x, cell_size_y

Cell resolution (from res)

n_na

Number of NA cells in the landscape

Metrics are identified by their abbreviation throughout. To see what an abbreviation stands for, look it up with list_lsm; evaluate_metrics reports the full names in its ranking table, and plot_metrics can use them as facet labels via metric_labels = "name".

The last five columns record each landscape's geometry so it stays attached to the metrics (e.g. through write_csv); they are used for geometry-mismatch checks and are never used as model predictors.

References

Hesselbarth, M.H.K., Sciaini, M., With, K.A., Wiegand, K., & Nowosad, J. (2019). landscapemetrics: an open-source R tool to calculate landscape metrics. *Ecography*, 42(10), 1648-1657. doi:10.1111/ecog.04617

See also

plot_metrics, list_lsm for the available metrics and their full names

Other metrics: evaluate_metrics(), print.metrics_evaluation()

Examples

# Calculate all landscape-level metrics for a single landscape
landscape <- create_landscape(pattern = "labyrinth")
metrics <- calculate_metrics(landscape)

# Calculate specific metrics for multiple landscapes
landscapes <- create_landscapes(n = 10, patterns = "spots")
#> ✔ Successfully generated all 10 training landscapes
metrics <- calculate_metrics(
  landscapes,
  metrics = c("ai", "lsi"),
  level = "landscape"
)