Compute internal metrics for spatial data — getSpatialInternalMetrics • poem

A generic function to compute a selection of internal clustering evaluation metrics for spatial data. It can be applied to raw components (labels, location) or directly to a SpatialExperiment object.

Usage

getSpatialInternalMetrics(
  object = NULL,
  labels,
  location = NULL,
  k = 6,
  level = "class",
  metrics = c("CHAOS", "PAS", "ELSA"),
  ...
)

# S4 method for class 'missing'
getSpatialInternalMetrics(
  object = NULL,
  labels,
  location = NULL,
  k = 6,
  level = "class",
  metrics = c("CHAOS", "PAS", "ELSA"),
  ...
)

# S4 method for class 'SpatialExperiment'
getSpatialInternalMetrics(
  object = NULL,
  labels,
  location = NULL,
  k = 6,
  level = "class",
  metrics = c("CHAOS", "PAS", "ELSA"),
  ...
)

Arguments

object: The main input. Can be a SpatialExperiment object or missing (when using labels, and location directly).
labels: When object is missing: a vector containing the labels of the predicted clusters. Must be a vector of characters, integers, numerics, or a factor, but not a list. When object is a SpatialExperiment object: the column name in colData(object) containing the labels.
location: A numerical matrix containing the location information, with rows as samples and columns as location dimensions.
k: The size of the spatial neighborhood to look at for each spot. This is used for calculating PAS and ELSA scores.
level: The level to calculate the metrics. Options include "element", "class" and "dataset".
metrics: The metrics to compute. See details.
...: Optional params for PAS().

Value

A data.frame of metrics.

Details

The allowed values for metrics depend on the value of level:

If level = "element", the allowed metrics are: "PAS", "ELSA".
If level = "class", the allowed metrics are: "CHAOS", "PAS", "ELSA".
If level = "dataset", the allowed metrics are:
- "PAS": Proportion of abnormal spots (PAS score)
- "ELSA": Entropy-based Local indicator of Spatial Association (ELSA score)
- "CHAOS": Spatial Chaos Score.
- "MPC": Modified partition coefficient
- "PC": Partition coefficient
- "PE": Partition entropy

Examples

# Example with individual components
data(sp_toys)
data <- sp_toys
getSpatialInternalMetrics(labels=data$label, location=data[,c("x", "y")], 
                          k=6, level="class")
#> the specified variable is considered as categorical...
#>   class         PAS    ELSA.Ea   ELSA.Ec       ELSA     CHAOS
#> 1     1 0.008333333 0.07106927 0.1940084 0.06040663 0.2305866
#> 2     2 0.008333333 0.06918805 0.1944579 0.05840334 0.2305866

# Example with SpatialExperiment object
se_object <- SpatialExperiment::SpatialExperiment(assays=matrix(NA, 
                                             ncol = nrow(data[,c("x", "y")]), 
                                             nrow = ncol(data[,c("x", "y")])), 
                               spatialCoords=as.matrix(data[,c("x", "y")]))
SummarizedExperiment::colData(se_object) <- 
cbind(SummarizedExperiment::colData(se_object), 
      data.frame(label=data$label))
getSpatialInternalMetrics(object=se_object, labels="label", k=6, 
                          level="class")
#> the specified variable is considered as categorical...
#>   class         PAS    ELSA.Ea   ELSA.Ec       ELSA     CHAOS
#> 1     1 0.008333333 0.07106927 0.1940084 0.06040663 0.2305866
#> 2     2 0.008333333 0.06918805 0.1944579 0.05840334 0.2305866