Package index • poem

Main functions

All metrics can be retrieved via the following six main wrapper functions.

getEmbeddingMetrics(): Compute embedding-based metrics

getGraphMetrics(): Compute graph-based metrics

getPartitionMetrics(): Compute partition-based metrics

getFuzzyPartitionMetrics(): Compute external metrics for fuzzy clusterings

getSpatialExternalMetrics(): Calculate Spatial External Metrics

getSpatialInternalMetrics(): Compute internal metrics for spatial data

Functions for individual metrics

These are some useful functions for calculating individual metrics.

Embedding metrics

CDbw(): Calculate CDbw index

dbcv(): Calculate DBCV Metric

silhouetteWidths(): silhouetteWidths

Fuzzy metrics

fuzzyHardMetrics(): Compute fuzzy-hard versions of pair-sorting partition metrics

fuzzyPartitionMetrics(): Compute fuzzy-fuzzy versions of pair-sorting partition metrics

Internal metrics for spatial domain detection

PAS(): Calculate PAS score

CHAOS(): Calculate CHAOS score

ELSA(): Calculate ELSA scores

External metrics for spatial domain detection

spatialARI(): Spatially aware ARI from Yan, Yinqiao, et al. (2025).

Element-wise external metrics for spatial domain detection, particularly useful for visualizations.

fuzzyHardSpotConcordance(): Per-element maximal concordance between a hard and a fuzzy partition

fuzzySpotConcordance(): Per-element concordance between two fuzzy partitions

getPairConcordance(): Per-element pair concordance score

getNeighboringPairConcordance(): Per-element local concordance between a clustering and a ground truth

Datasets

The package provides several datasets for demonstration purposes.

noisy_moon: The noisy moon dataset

sp_toys: Toy examples of spatial data

toyExamples: Toy embedding examples

metric_info: Metrics Information

Other utility functions

mockData(): Generate mock multidimensional data

matchSets(): Match two partitions using Hungarian algorithm

knnComposition(): Compute neighborhood composition

getFuzzyLabel(): Get fuzzy representation of labels

findSpatialKNN(): Find the k nearest spatial neighbors

emb2knn(): Computes k nearest neighbors from embedding

emb2snn(): Computes shared nearest neighbors from embedding