I have a data frame of 36 experimentees, who each executed 500 Trials in an experiment and because of this size i would like to make a function of some sort for my problem.

An exemplary data frame could look like this:

data <- data.frame(ID = c(1, 1, 1, 2, 2, 2, 3, 3, 3),
               trial = c(1, 2, 3, 1, 2, 3, 1, 2, 3),
               value = c(6, 8, 2, 5, 7, 2, 8, 9, 2))

"ID" is the identifier for the experimentee, "trial" is obviously the trial which was executed and "value" is a numeric value (in the exemplary data frame just random numbers).

I want to create a new column and for every experimentee I want the value of a trial subtracted by the value of the subsequent trial.

So value of trial 1 minus value of trial 2; value of trial 2 minus value of trial 3 and so on..

basically "value(trial(n)) - value(trial(n 1))".

It doesn´t seem to be a very difficult problem and I still can´t figure it out. Any help would be great, I am fairly new to R.

CodePudding user response：

You can group_by() the experimenteer "ID" and then calculate your value using the dplyr::lead() function, i.e.

   library(dplyr)
   data %>% 
    group_by(ID) %>% 
    mutate(new_var = value - dplyr::lead(value))

Output:

# A tibble: 9 x 4
# Groups:   ID [3]
     ID trial value new_var
  <dbl> <dbl> <dbl>   <dbl>
1     1     1     6      -2
2     1     2     8       6
3     1     3     2      NA
4     2     1     5      -2
5     2     2     7       5
6     2     3     2      NA
7     3     1     8      -1
8     3     2     9       7
9     3     3     2      NA

CodePudding user response：

data.table option using shift:

library(data.table)
setDT(data)
data[, new_var := value-shift(value, type ="lead"), by = ID][]

Output:

   ID trial value new_var
1:  1     1     6      -2
2:  1     2     8       6
3:  1     3     2      NA
4:  2     1     5      -2
5:  2     2     7       5
6:  2     3     2      NA
7:  3     1     8      -1
8:  3     2     9       7
9:  3     3     2      NA