Utils: Utilities

In matloff/qeML: Quick and Easy Machine Learning Tools

Utilities

Description

Miscellaneous functions, some used mainly internally in the package, but of possible use externally.

Usage

buildQEcall(qeFtnName,dataName,yName=NULL,opts=NULL,holdout=NULL,
    holdoutArg=TRUE) 
evalr(toexec) 
newDFRow(dta,yName,x,dtaRowNum=1)
replicMeansMatrix(nReps,cmd,nCols=NULL)
Data(datasetName)
wideToLongWithTime(data,timeColName,timeColPresent=TRUE,
   timeColSeq=c(1,1),grpColName=NULL,valueColName=NULL)

Arguments

data	Data frame or equivalent. Other than a "time" column, all columns must be numeric, one column per group (see below).
qeFtnName	Quoted name of a qeML predictive function.
dataName	Quoted name of a data frame.
yName	Quoted name of a column to be predicted.
opts	Non-default arguments for the function specified in qeFtnName.
holdout	Size of holdout set, if any.
holdoutArg	A value TRUE means the function specified in qeFtnName has an argument 'holdout'.
toexec	Quoted string containing an R function call.
nReps	Number of replications.
cmd	Quoted string containing an R function call. If multiple statements, enclose with braces.
nCols	Number of columns for output.
dta	A data frame.
x	An R list specifying fields to be set.
dtaRowNum	Row number in 'dta' to be used as a basis.
datasetName	Quoted string of dataset to be loaded.

Details

The function qeFtnName does what its name implies: It assembles a string consisting of a qeML function call. Typically the latter is then executed via evalr. See for instance the source code of qeLeaveOut1Var.

R's generic predict function generally required that the input rows match the original training data in name and class. The newDFRow function can be used to construct such a row.

The function replicMeansMatrix will eventually replace regtools::replicMeans. It runs the specified code many times, with the code assumed to have some random component such as in simulation or in investigation of a random algorithm.

The function Data is a convenience function that combines calls to data and str.

The function wideToLongWithTime plot "Y" versus "X" for different groups, one per column in data other than "X", if present. "X" is typically time but can be any ordered numeric quantity. If "X" is not present, the arguments specify how the user wants one created.

Author(s)

Norm Matloff

Examples

# function to list all the objects loaded by the specified package
lsp <- function(pkg) {
   cmd <- paste('ls(package:',pkg,')')
   evalr(cmd)
}
lsp('regtools')
# outputs
#  [1] "clusterApply"        "clusterApplyLB"      "clusterCall"        
#  [4] "clusterEvalQ"        "clusterExport"       "clusterMap"         
# ...

# mean of scalar quantity
replicMeansMatrix(1000,'rnorm(1)^2')
# mean of vector quantity
replicMeansMatrix(1000,'c(rnorm(1),rnorm(1)^2)',2)
# mean of matrix quantity
replicMeansMatrix(1000,
   '{z1=rnorm(1); z2=rnorm(1); x=z1; y=z1+z2; rbind(c(x,x^2),c(y,y^2))}',2) 

data(currency)
zLong <- wideToLongWithTime(currency,'weeknum',timeColPresent=FALSE,
   grpColName='country',valueColName='rate')

matloff/qeML documentation built on Dec. 15, 2024, 10:15 a.m.