Question

Say I have loaded a csv file into R with two columns (column A and column B say) with real value entries. Call the dataframe df. Is there away of speeding up the following code:

dfm <- df[floor(A) = x & floor(B) = y,]
x <- 2
y <- 2
dfm

I am hoping there will be something akin to function e.g.

dfm <- function(x,y) {df[floor(A) = x & floor(B) = y,]}

so that I can type

Any help much appreciated.

Answer 1

The way that’s written right now won’t work for a few reasons:

1. You need to assign values to x and y before you assign dfm. In other words, the lines x <- 2 and y <- 2 must come before the dfm <- ... line.
2. R doesn’t know what A and B are, even if you put them inside the brackets of the dataframe that contains them. You need to write df$A and df$B.
3. = is the assignment operator, but you’re looking for the logical operator ==. Right now your code is saying “Assign the value x to floor(A)” (which doesn’t really make sense). You want to tell it “Only choose rows where floor(A) equals x”, or floor(A)==x.

So what you want is:

dfm.create <- function(x,y) {df[floor(df$A)==x & floor(df$B)==y,]}
dfm <- dfm.create(2,2)

Note that if you want the dataframe to be called dfm, you don’t want to name the function dfm, or you will have to erase the function to make the dataframe.