getis: getis

In RichardLemoine/LSTtools: LSTtools: An R package to process thermal data derived from Landsat and MODIS images

Description

Compute a hot and cold spots analysis employing Getis-Ord Gi* statistic and the False Discovery Rate (FDR) correction.

Usage

getis(x, dist = NULL, p = 0.05)

Arguments

x	Raster layer with values to analyze.
dist	Distance for local neighborhood size from each cell.
p	P-value to reject the null hypothesis.

Details

Compute a hot and cold spots analysis based on the Getis-Ord Gi* statistic and the False Discovery Rate (FDR) correction. Center cell is included in the local neighborhood size and the limit is the distance specified by dist employing queen's case. False Discovery Rate (FDR) correction is applied to potentially reduce the critical p-value thresholds in order to account for multiple testing and spatial dependency. Possible p values to reject the null hypothesis are 0.1, 0.05, 0.01, 0.001 and 0.0001.

Value

A shapefile with the resulting z-scores, FDR values and label of cluster (i.e., hot spot, cold spot or no sig).

References

Benjamini, Y., and Hochberg, Y. (1995). Controlling the false discovery rate: a practical and powerful approach to multiple testing. Journal of the Royal Statistical Society Series B, 57, 289-300.

Getis, A. and Ord, J. K. (1996). Local spatial statistics: an overview. In P. Longley and M. Batty (eds) Spatial analysis: modelling in a GIS environment (Cambridge: Geoinformation International), 261-277.

Examples

## Not run: # For an LST layer derived from Landsat 8
spots <- getis(lstl8, dist = 70, p = 0.05)
## End(Not run)

RichardLemoine/LSTtools documentation built on Oct. 19, 2020, 12:50 a.m.