如何考虑R中每个观察值的下一行和上一行

Question

I need to create 2 columns: PRETARGET and TARGET based on several conditions.

To create PRETARGET , for each row of my data (for each participant PPT and trial TRIAL ) I need to check that the CURRENT_ID is associated with a value of 0 in the column CanBePretarget , and that the following row is the value of CURRENT_ID + 1. If these conditions are fulfil, then I would like to have a value of 0, if they are not fulfil a value of 1.

To create TARGET , for each row of my data (for each participant PPT and trial TRIAL ) I need to check that the CURRENT_ID is associated with a value of 0 in the column CanBeTarget , and that the previous row is the value of CURRENT_ID - 1. If these conditions are fulfil, then I would like to have a value of 0, if they are not fulfil a value of 1.

In addition, if the result in PRETARGET is 1, then the value of the next row in TARGET should also be 1.

I have added the desired output in the following example.

I was thinking to use for loops and ifelse statements, but I am not sure how to consider the following/previous row of each observation.

PPT   TRIAL PREVIOUS_ID CURRENT_ID  NEXT_ID CURRENT_INDEX CanBePretarget CanBeTarget    PRETARGET   TARGET
ppt01 11    2           3           4       3             0              0              0           1
ppt01 11    3           4           3       4             1              0              1           0
ppt01 11    4           5           6       8             0              0              1           1
ppt01 11    6           7           8       10            0              0              1           1
ppt01 11    7           10          11      18            0              1              0           1
ppt01 11    10          11          12      19            0              0              0           0
ppt01 11    11          12          14      20            1              0              1           0
ppt01 12    1           2           1       2             1              0              1           1
ppt01 12    2           3           4       5             0              0              1           1
ppt01 12    5           6           6       8             0              0              0           1
ppt01 12    6           7           7       10            0              0              0           0
ppt01 12    7           8           9       12            0              0              0           0
ppt01 12    8           9           9       13            0              0              0           0
ppt01 12    9           10          11      16            0              0              0           0
ppt01 12    10          11          11      17            0              0              0           0
ppt01 13    1           2           2       2             1              0              1           1
ppt01 13    3           3           3       10            0              0              1           1
ppt01 13    4           5           6       13            0              0              0           1
ppt01 13    5           6           7       14            0              0              1           0
ppt01 13    9           9           10      19            0              0              0           1
ppt01 13    9           10          10      20            0              0              0           0
ppt01 13    10          11          12      22            0              0              0           0
ppt01 13    11          12          12      23            0              0              1           0
ppt01 14    10          11          11      15            0              0              0           1
ppt01 14    11          12          12      17            0              0              1           0

Answer 1

This can be achieved by using dplyr

df.new <- df %>% 
          mutate(PRETARGET1 = abs(as.numeric(CanBePretarget == 0 & lead(CURRENT_ID, default = 0) == (CURRENT_ID + 1)) - 1)) %>% 
          group_by(PPT, TRIAL) %>% 
          mutate(TARGET1 = abs(as.numeric((CanBeTarget == 0 & lag(CURRENT_ID, default = 0) == (CURRENT_ID - 1)) ) -1),
                 TARGET1 = ifelse(lag(PRETARGET1, default = 0) == 1, 1, TARGET1))

To compare to your results, I created PRETARGET1 and TARGET1 .