IQD: Calculate the IQ distance at each location

In NorskRegnesentral/climModelEval: Climate model evaluation tools

Description

At each location, calculate IQ distance between the ecdf of a single data set of a climate model ('modelArray') and the ecdf of obervations ('obsArray'). Both input arrays are three-dimensional (longitude, latitude, time). The function IQDlocal() is accessed with two options: the crps_sample function (crps=TRUE) and the integrate funtion (crps=FALSE). The former is much faster and with no problems during the calculations. The integrate function is problematic due to the 'subdivisions' (the maximum number of subintervals), i.e. no convergence even though the number of subintervals is large.

Usage

IQD(modelArray, obsArray, crps, remove_indices = NULL)

Arguments

modelArray	Array with model data, 3d-array (longitude, latitude, time)
obsArray	Array with dataproduct data, 3d-array (longitude, latitude, time)
crps	If true, crps is used (strongly recomended), otherwise integrate is used
remove_indices	Which indices to remove. Defaults to NULL

Examples

IQD()

NorskRegnesentral/climModelEval documentation built on Jan. 16, 2020, 12:49 a.m.