scoreOrderComparison {OrderedList} | R Documentation |
Compute weighted similarity score for gene rankings determined via the chosen test statistics.
scoreOrderComparison(exprs1, labels1, paired1, exprs2, labels2, paired2, test.method = test.z, nn, bases, two.sided) scoreOrderComparisonBoth(exprs1, labels1, paired1, exprs2, labels2, paired2, test.method = test.z, nn, bases, two.sided)
exprs1,exprs2 |
gene expression matrices. |
labels1,labels2 |
class labels, one label per column in matrices. |
paired1,paired2 |
logical, whether samples are paired in classes. |
test.method |
a function computing one test statistics per row and taking
a matrix, a label vector and a logical for pairing as parameters.
Valid examples are test.fc , test.t and test.z . |
nn |
a vector of rank limits. The score is computed taking into
account ranks up to these limits only. One limit per entry in bases . |
bases |
a vector of bases used in weighted scores, is equal to exp(-alpha). The function can compute scores for several regularization parameters in one go. |
two.sided |
if TRUE both ends of the lists are taken into account, only
top ranks are considered otherwise. |
scoreOrderComparison
computes scores only for the direct
comparison. scoreOrderComparisonBoth
in addition computes
scores for reversed orderes, i.e., one of the rankings is reversed.
For each entry in bases
, thus for each regularization parameter alpha,
one score is returned in an array.
Claudio Lottaz