I am using RTextTools, which has a function to create container with following syntax:

create_container(matrix, labels, trainSize=NULL, testSize=NULL, virgin)

Its description states about virgin:

A logical (TRUE or FALSE) specifying whether to treat the classification data as virgin data or not.

What does "virgin data" mean?

  • $\begingroup$ The best docs ever written right there $\endgroup$ – wordsforthewise Nov 17 '17 at 21:08

When the virgin flag is set to FALSE, it indicates that all data in the training and testing sets have corresponding labels.

When the virgin flag is set to TRUE, it indicates that the testing set is unclassified data with no known true values.

  • 7
    $\begingroup$ "virgin" sure is a cryptic way to say unlabeled test set. $\endgroup$ – Marc Claesen Dec 27 '14 at 9:47

The answer by @pratik_m is correct.

As a historical note, this terminology comes from Laver et al. 2003. Although it's not stated explicitly, Laver et al. use 'virgin' to refer to a document that a) has no target value because b) it is out of sample. What makes things confusing is that in the paper (though not in RTextTools) the target value is not actually a label but a real valued position. Also, while all out-of-sample documents will lack a target value, not all documents lacking a target value must be out-of-sample. (Lowe 2008 shows that if the missing target values are treated as in-sample then the algorithm in the paper computes a correspondence analysis with a subset of row scores fixed).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.