oss.optSpline: Evaluate splines with multiple values of lambda

In newdale/onsoilsurvey: Making PDSM in Ontario Better

Evaluate splines with multiple values of lambda

Description

This function is a wrapper for the ea_spline function from the ithir package maintained by Brendan Malone and available on bitbucket. The wrapper allows the user to evaluate and plot equal area quadratic splines using multiple values of lambda.

Usage

oss.optSpline(
  obj,
  var.name,
  lam = 0.1,
  d = t(c(0, 5, 15, 30, 60, 100, 200)),
  vlow = 0,
  vhigh = 1000,
  dir = getwd(),
  save.plots = TRUE
)

Arguments

obj	Object of class "data.frame"
var.name	character; target variable name (must be a numeric variable)
lam	numeric; lambda the smoothing parameter
d	numeric; standard depths that are used to extract values from fitted spline i.e harmonising depths
vlow	numeric; smallest value of the target variable (smaller values will be replaced)
vhigh	numeric; highest value of the target variable (larger values will be replaced)
dir	character; location to store plots if the option is selected, defaults to current working directory
save.plots	logical; TRUE or FALSE, to generate plots or not

Value

Returns a list with five elements: harmonised data frame; are spline-estimated values of the target variable at standard depths obs.preds data frame; are observed values together with associated spline predictions for each profile at each depth. splineFitError matrix; rmse and rmseiqr between observed and predicted between observed profiles and associated fitted splines. var.1cm matrix; are spline-estimated values of the target variable using the 1 cm increments lamdf data frame; returns the optimal value of lambda for each profile

Examples

#Generate splines with various lambdas for soil profiles
data(profiles)
d=t(c(0,5,15,30,60,100))
lam=c(10, 1, 0.1, 0.01, 0.001)
oss.optSpline(obj=profiles,var.name="CEC",lam=lam,d=d,save.plots=FALSE)

newdale/onsoilsurvey documentation built on Jan. 5, 2024, 1:35 a.m.