[R] weird problem - R is not finding the data for the factor level present in the data

[Dimitri Liakhovitski]

Sorry for no code - but it's a more of a general question.
I have read in a data frame ("|"-delimited, .txt).
daily<-read.table(file="filename.txt",sep="|",header=T)

One of the variables is a factor with 110 levels:
>str(daily$dma_id)
 Factor w/ 110 levels "500","501","503",...

108 levels of this factor happen to be numbers "500", "501", "503",
... "880","881"
But the last 2 levels are strings:

>levels(daily$dma_id)[109:110]
[1] "OH1054" "PA2207"

I checked in the raw data file (.txt) that there are no spaces in
these last 2 levels, that there is nothing weird. There is nothing.

When I do the following with any level of that factor, I get the data
I am expecting:
>daily$dma_id[daily$dma_id %in% levels(daily$dma_id)[108]]
[1] "881" "881" "881" - etc.

But when I do the same for the last two levels, I get nothing, for example:
>daily$dma_id[daily$dma_id %in% "OH1054"]  # - or:
>daily$dma_id[daily$dma_id %in% levels(daily$dma_id)[109]]
factor(0)

As a result I can't rename those levels.

It's not a real problem - I can do replace in the raw txt file. But
still, why is it happening?

Thank you!



-- 
Dimitri Liakhovitski
Ninah Consulting
www.ninah.com