Perform a Single Classification
runTest.Rd
For a data set of features and samples, the classification process is run. It consists of data transformation, feature selection, classifier training and testing.
Usage
# S4 method for matrix
runTest(measurementsTrain, outcomeTrain, measurementsTest, outcomeTest, ...)
# S4 method for DataFrame
runTest(
measurementsTrain,
outcomeTrain,
measurementsTest,
outcomeTest,
crossValParams = CrossValParams(),
modellingParams = ModellingParams(),
characteristics = S4Vectors::DataFrame(),
...,
verbose = 1,
.iteration = NULL
)
# S4 method for MultiAssayExperiment
runTest(measurementsTrain, measurementsTest, outcomeColumns, ...)
Arguments
- measurementsTrain
Either a
matrix
,DataFrame
orMultiAssayExperiment
containing the training data. For amatrix
orDataFrame
, the rows are samples, and the columns are features.- ...
Variables not used by the
matrix
nor theMultiAssayExperiment
method which are passed into and used by theDataFrame
method or passed onwards toprepareData
.- outcomeTrain
Either a factor vector of classes, a
Surv
object, or a character string, or vector of such strings, containing column name(s) of column(s) containing either classes or time and event information about survival. If column names of survival information, time must be in first column and event status in the second.- measurementsTest
Same data type as
measurementsTrain
, but only the test samples.- outcomeTest
Same data type as
outcomeTrain
, but for only the test samples.- crossValParams
An object of class
CrossValParams
, specifying the kind of cross-validation to be done, if nested cross-validation is used to tune any parameters.- modellingParams
An object of class
ModellingParams
, specifying the class rebalancing, transformation (if any), feature selection (if any), training and prediction to be done on the data set.- characteristics
A
DataFrame
describing the characteristics of the classification used. First column must be named"charateristic"
and second column must be named"value"
. Useful for automated plot annotation by plotting functions within this package. Transformation, selection and prediction functions provided by this package will cause the characteristics to be automatically determined and this can be left blank.- verbose
Default: 1. A number between 0 and 3 for the amount of progress messages to give. A higher number will produce more messages as more lower-level functions print messages.
- .iteration
Not to be set by a user. This value is used to keep track of the cross-validation iteration, if called by
runTests
.- outcomeColumns
If
measurementsTrain
is aMultiAssayExperiment
, the names of the column (class) or columns (survival) in the table extracted bycolData(data)
that contain(s) the samples' outcome to use for prediction.
Value
If called directly by the user rather than being used internally by
runTests
, a ClassifyResult
object. Otherwise a
list of different aspects of the result which is passed back to runTests
.
Details
This function only performs one classification and prediction. See
runTests
for a driver function that enables a number of
different cross-validation schemes to be applied and uses this function to
perform each iteration.
Examples
#if(require(sparsediscrim))
#{
data(asthma)
tuneList <- list(nFeatures = seq(5, 25, 5), performanceType = "Balanced Error")
selectParams <- SelectParams("limma", tuneParams = tuneList)
modellingParams <- ModellingParams(selectParams = selectParams)
trainIndices <- seq(1, nrow(measurements), 2)
testIndices <- seq(2, nrow(measurements), 2)
runTest(measurements[trainIndices, ], classes[trainIndices],
measurements[testIndices, ], classes[testIndices], modellingParams = modellingParams)
#> An object of class 'ClassifyResult'.
#> Characteristics:
#> characteristic value
#> topN 5
#> Balanced Error 0.147058823529412
#> Selection Name Moderated t-test
#> Classifier Name Diagonal LDA
#> Cross-validation Independent Set
#> Features: List of length 1 of feature identifiers.
#> Predictions: A data frame of 95 rows.
#> Performance Measures: None calculated yet.
#}