TrainTestResult: Class to handle the results of Training-Testing evaluation...
In elqumsan/RNAseqMVA: Application of Multivariate Analysis methods to RNA-seq Data

Description Usage Arguments Value Author(s)

Each an object of this class contains the primary results of a training-testing evaluation of a classifier

TrainTestResult(
  dataset,
  classifier,
  iteration,
  trainPredictedClasses,
  testPredictedClasses
)

`dataset`	the dataset on which the classifier ran
`classifier`	the classifier used to produce this result
`iteration`	is the number of resampling for sake of much more precision to evaluate used classifier.
`trainPredictedClasses`	are the classes that are predicted based on training the classifier with trainIndices.
`testPredicatedClasses`	are the classes that are predicated based on training an classifier with testIndices.
`call`	the function call used to produce this result

an object of class TrainTestResult containing the following attributes

ID: is that recountID for the test dataset.
classLabels: are labels (classes) for each sample in the targeted dataset.
dataType: are the feature type (preprocessing created data) to analyize the imapct of pre-processing into efficiency of classifer, e.g. (filtered, norm, log2norm, log2norm-Pcs, ....)
parameters: are all general parameter joind with each dataset from methods init_recount_id().
classifier: is the classifier used to produce such results.
iteration: is number of resampling to individuals in order to paly it with classifier.
trainIndices: are the whole indeices for the trainset to be used as the trainset for the classifier ran.
testIndices: are the whole indices for the testset to be paased to classifier as the tesset part.
trainPredictedClasses: all the classes that are predicted based on training the classifier with trainset.
testPredictedClasses: all the classes that are predicted based on training the classifier with testset.
test.contingency: is (testing error) that is a table which contains in raw the actual classes on the testing set and in column predicted classes from classifier on the testing set.wherein that in diagolan is the correct classification and out of diagonal is miscalssifcation error rate for the testing set.
testing.errors: are the class labels that are existed in the actuall classes for the testset and which are not existed in the predicted classes from the trained classifier.
testing.error.nb: is the sume of the testing.errors.
testing.error.rate: is the training.error.nb divided by whole counts of the testing errors.
train.contingency: is (learning error) the table contain in the raw actual class and in column predicted class and in diagonal is the corect classification and out of diagonal is misclassification errors.
training.errors: are the class labels that are founded in the actuall classes for the trainset and which are not existed in the predicted classes from the trained classifier.
training.error.nb: is the sum of the training.errors
training.error.rate: is the training.error.nb divided by overall counts of the training errors.
trainProportion: the ratio of the trainset from all the dataset.
trainSize: is volume of the train size from all the dataset it is computed by multiple number of individuals in the trainProportion magnitude.
testSize: the test size "remained size" from all dataset after we remove the trainSize from targeted dataset