用另一个因素替换一个因素（在数据框内）

Question

I have two factors and they do not have the same number of levels, but I would like to use one factor to replace values in the other factor inside a data-frame based on the name and the order of the factors.

My data looks like this,

x <- factor(c("one", "two", "three", "two", "three"))
y <- factor(c(NA, "foo", NA, "bar", NA))

(df <- data.frame(x, y))

      x    y
1   one <NA>
2   two  foo
3 three <NA>
4   two  bar
5 three <NA>

and this is where I would like to end up,

      x    y     z
1   one <NA>   one
2   two  foo   foo
3 three <NA> three
4   two  bar   bar
5 three <NA> three

Should I convert the factors to a character vector?

Answer 1

you could use levels(z) <- c(levels(y), levels(x)) so that z has the required levels, however the underlying integer values may not relate correctly. You are probably better off assigning to z using as.character and then converting to factor.

eg

df$z <- as.factor( ifelse(is.na(df$y), as.character(df$x), as.character(df$y)) )