[R] ifelse for creating discriminating variable based on two conditions

[Andreas Nord]

Dear list,

Apologies for a likely naive question.


I am trying to create a discriminating dummy variable using 'ifelse' based on conditions in two variables.


Specifically, I want to assign levels in a new factor as '0' or '1' based on a user-defined cut off. I.e. something similar to:

  >data1<-data.frame(molecule=runif(30,min=0,max=1e3))

>data1$newcol<-ifelse(data1$molecule>2*sd(data1$molecule),1,0)


Which is all straightforward.


But how do I go on to assign values in variable 'molecule'based on the same cut off, but separately for each level of a second variable, in this case the factor 'fruit' with three levels. That is, how do I derive fruit-specific cut-offs using a data frame with the general structure of that below?

>data2<-data.frame(molecule=runif(30,min=0,max=1e3),fruit=factor(rep(c('apple','pear','orange'),10)))


Many thanks in advance!



	[[alternative HTML version deleted]]