Skip to content
Prev 178355 / 398506 Next

re moving entries from one vector that are in another

onyourmark 
I have various objects defined but I am trying to remove a set of elements in
one vector from another and using the loops at the end of this post but I am
getting the error at the very end of this post.
num [1:923, 1:923] 1 -0.00371 -0.00102 -0.00204 -0.00102 ...
 - attr(*, "dimnames")=List of 2
  ..$ : chr [1:923] "a4.1" "abdomen.2" "abdomimal.3" "abdominal.4" ...
  ..$ : chr [1:923] "a4.1" "abdomen.2" "abdomimal.3" "abdominal.4" ...
int [1:2129, 1:2] 1 399 653 2 3 600 4 5 271 870 ...
 - attr(*, "dimnames")=List of 2
  ..$ : chr [1:2129] "a4.1" "hirschsprung.399" "peritoneal.653" "abdomen.2"
...
  ..$ : chr [1:2] "row" "col"
int [1:1206, 1:2] 399 653 600 271 870 185 298 620 119 162 ...
 - attr(*, "dimnames")=List of 2
  ..$ : chr [1:1206] "hirschsprung.399" "peritoneal.653" "occult.600"
"enteroclysis.271" ...
  ..$ : chr [1:2] "row" "col"


#Trying to delete all variables in my correlation matrix x that have
correlation greater than .6 
answer2<-which(((x >.6) | (x<(-.6))), arr.ind = TRUE)

#also need to delete the diagonal entries of x  (where a var is correlated
with itself) because my goal is to variables from mydataN but a variable
being correlated with itself is not a reason to drop it:
answer3<-answer2
answer3 <- answer2[answer2[,1]!=answer2[,2],]

#so now the second row of answer3 is a list of highly correlated variables.

#now take the entire 2nd column answer3 :in other words we want a list of 

correlated variables that we are going to eliminate.


uniqueFromColumn2ofAnswer2=unique(answer3[,2])
int [1:561] 1 3 5 10 12 13 15 17 18 19 ...
#now create a holder for mydataN minus those columns which have "bad" or
highly correlated variables. 

mydataNMinusHighCorForAll <- mydataN

#and now go ahead and take them out:

for (i in uniqueFromColumn2ofAnswer2) {for (j in
mydataN[0,]){if(i==j){mydataNMinusHighCorForAll[,-j]}}}

Error in if (i == j) { : argument is of length zero