Thank you for your kind responses. I have a dataframe and I need to apply a formula over a columns depending of the value of other column:
data <- data.frame(figure = c("square","square","circle","square","circle"),
diameter =c(NA,NA,21,NA,12),
side=c(32,27,NA,51,NA))
What I need is to calculate the area according the square or circle formula (square = side * side, circle= pi*(diameter/2)^2) in a new column:
data$area <- 0
data$area[data$figure == "square"] <- data$side*data$side
data$area[data$figure == "circle"] <- pi*(data$diameter/2)^2
But when I calculate this way, I get the last formula and the following error:
The number of items to replace is not a multiple of the length of the replacement.
I suposse this could be related with column selection in brackets. Any help will be appreciated.
>Solution :
We could use ifelse
data$area <- with(data, ifelse(figure == "square", side^2, pi*(diameter/2)^2))
The lhs and rhs should have the same length. In the OP’s code, the lhs is of different length than the rhs (i.e. full column values are used in rhs) i.e.
data$area[data$figure == "square"] <- (data$side*data$side)[data$figure == "square"]
data$area[data$figure == "circle"] <- (pi*(data$diameter/2)^2)[data$figure == "circle"]
-output
> data
figure diameter side area
1 square NA 32 1024.0000
2 square NA 27 729.0000
3 circle 21 NA 346.3606
4 square NA 51 2601.0000
5 circle 12 NA 113.0973