model_ols_gpm: Model with OLS and GPM Components

In alexmitrani/humblr: Provides a few humble tools for R.

Model with OLS and GPM Components

Description

produces a model with Ordinary Least Squares (OLS) and Gaussian Process Model (GPM) components, based on input data provided in an Excel spreadsheet.

Usage

model_ols_gpm(
  myinputspreadsheet = NULL,
  myinputdatasheet = NULL,
  mypredictiondf = NULL,
  mytag = NULL,
  mypredictionsheet = NULL,
  myresponse_ols = NULL,
  myresponse_gpm = "residuals",
  myterms = NULL,
  sigmafrom = 0.01,
  sigmato = 2,
  sigmaby = 0.01,
  myfolds = 10,
  myseed1 = 123,
  myseed2 = 456,
  mygraphsize = 1000
)

Arguments

myinputspreadsheet	is the name of the spreadsheet with the input data.
myinputdatasheet	the name of the sheet with the input data.
mypredictiondf	(optional) a dataframe to be used as a basis for predictions
mytag	a character string that will be used in the output file names. #'
mypredictionsheet	(optional) the name of a sheet containing test data to which the model should be applied.
myresponse_ols	the name of the dependent variable for the OLS model
myresponse_gpm	the name of the dependent variable for the GPM model. This will normally be the residuals of the OLS model: "residuals".
myterms	a character vector containing the names of the independent variables.
sigmafrom	the lower limit of the test range for sigma
sigmato	the upper limit of the test range for sigma
sigmaby	the increment of the test values for sigma
myfolds	the number of folds for k-fold cross-validation
myseed1	the first random number seed
myseed2	the second random number seed
mygraphsize	the pixel dimension for the output graphs

Details

The spreadsheet requires at least one sheet, that must contain the dependent and independent variables.

Examples

myinputfile <- system.file("extdata", "model_ols_gpm_test.xlsx", package = "humblr")
mytestmodel <- model_ols_gpm(myinputspreadsheet = myinputfile, myinputdatasheet = "data", mypredictionsheet = "prediction_tests", mytag = "model_ols_gpm_test", myresponse_ols = "depvar", myresponse_gpm = "residuals", myterms = c("var2",	"var3",	"var4",	"var5",	"var6",	"var7",	"var8",	"var9",	"var10"), sigmafrom = 1, sigmato= 10, sigmaby = 1, myfolds = 10, myseed1 = 123, myseed2 = 456, mygraphsize = 1000)

alexmitrani/humblr documentation built on April 4, 2022, 8:29 a.m.