I am trying to reorder a factor data type so that when I plot stats
associated with the factor, the ordering makes sense.
For instance, if I have a factor entered as follows ...
A = as.factor(c("1", "10", "3", "3", "10", "10"))
levels(A)
... the ordering does not really make sense (assuming I want the factor
ordered by integer value), but I understand that this "mis-ordering" is
because the ordering is based on a character string data type and not on an
integer data type. Because I run into this problem frequently, I wrote a
small function to fix this:
reorder_factor = function(x, x_sum, decreasing=FALSE){
factor(as.character(x), levels=levels(x)[order(x_sum, decreasing=decreasing)
])
}
I can then run the following code to fix the problem:
A = reorder_factor(x=A, x_sum=as.numeric(levels(A)), decreasing=FALSE)
levels(A)
... and now I have correctly ordered integers. Perhaps not the most elegant
solution, but it worked for my purposes. Now I have a more complicated
problem and I need help. Assuming the following factor:
B = as.factor(c("Engine 1", "Engine 10", "Ladder 3", "Engine 3", "Ladder 10",
"Engine 10"))
levels(B)
I would like the factor ordered first by the proceeding unit type and then
ordered by the following integer. In this case, I would like to see this
order: Engine 1, Engine 3, Engine 10, Ladder 3, Ladder 10. I have tried
many different ways of separating out the unit type from the number, but am
having trouble figuring out a good way of achieving this factor order. For
such a small example, I could obviously manually change the order, but I am
dealing with much larger datasets with many unit types and up to 20
different numbers for each unit type. Having an automated way of ordering
these units would be a huge help. Thanks in advance for any help you can
provide.
--Markus Weisner
[[alternative HTML version deleted]]
______________________________________________
[email protected] mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
