findmutualinfos {rWMBAT} | R Documentation |
Given a set of data (many cases, each with values for many variables) and an additional value stored in the vector class, it finds MI described below in "Value."
findmutualinfos(data, class)
data |
a number of cases (in rows), each with a measurement for a group of variables (in columns). The data should be discredited into integers 1 through k. The columns are considered variables V1, V2, ... |
class |
a column vector of length "cases" with integer values 1,2..., an additional measurement of class C |
mi_vc |
a row double vector whose ith value is MI(Vi,C) |
mi_vv |
double symmetric matrix with values MI(Vi,Vj) |
mi_vc_v |
double nonsym matrix with values MI(Vi,C;Vj) |
CALLED FUNCTIONS MutualInfo(Calculate Mutual Information Of Two Variables) CondMutualInfo(Calculate Mutual Information of Two Variables Conditioned On a Third)
Karl Kuschner, Qian Si and William Cooke, College of William and Mary, Dept. of Physics, 2009
http://kwkusc.people.wm.edu/dissertation/dissertation.htm
data(traingrpbin, traingrpclass) # load example input data from package result <- findmutualinfos(traingrpbin, traingrpclass)