compeval: Provides a range of statistics for evaluating how components...

In johnlpearce/sommix: Tools to Apply Self-Organizing Maps

Description

Calculates a range of statistics for components (i.e., variables) of the training data in order to better understand relationships with the SOM model.

Usage

compeval(som_obj, labsize = 1, colormod = NULL)

Arguments

som_obj	an object returned from sommix_summ
labsize	sets label size on plots
colormod	sets colors for plots

Details

Understanding how well components relate to the SOM model can provide insights into variable importance. Here correlations, root means square error, and mean absolute error are calculated using observed vs predicted values from the SOM model. Measures of spatial autocorrelation are provided to enhance understanding of spatial patterning among components on the SOM.

Value

a list containing component evaluation statistics

• Component_Fit a dataframe of model fit statistics for each component/variable

• Component_Map a dataframe of Moran's I statistics for each component/variable

johnlpearce/sommix documentation built on Jan. 7, 2021, 11:38 p.m.