Perform a conditional calculation on every row value, in a single col, with the all the row vals after current in a data.table

[monty13]

Hello,

I appreciate that the title might be confusing so I will try to explain more here.

So, consider the following data.table:

sampleDT <- data.table(id = c(1,2,3,4,5),
                        dist = c(111.0200,124.0074,124.1337,128.0353,132.0303),
                        score = c(2985048,21641962,2627225,10382804,4338643)
)

I would like to subtract all the distances (column: dist) with the next ones in the array and make a conditional check for the resulted value. Pseudocode below to demo the logic:

if(dist[1]-dist[2] == xval) then sampleDT[, point := "Keep"] else continue,
if(dist[1]-dist[3] == xval) then sampleDT[, point := "Keep"] else continue,
if(dist[1]-dist[4] == xval) then sampleDT[, point := "Keep"] else continue,
if(dist[1]-dist[5] == xval) then sampleDT[, point := "Keep"] else continue,
if(dist[2]-dist[3] == xval) then sampleDT[, point := "Keep"] else continue,
if(dist[2]-dist[4] == xval) then sampleDT[, point := "Keep"] else continue,
if(dist[2]-dist[5] == xval) then sampleDT[, point := "Keep"] else continue,
.
.
.
if(dist[4]-dist[5] == xval) then sampleDT[, point := "Keep"] else continue

Visual representation of the hops I want to be performed.

Is something like this feasible?

[martin.R]

I'm not sure what you expect the output to look like, but you can use shift():
Fast lead/lag for vectors and lists — shift • data.table (rdatatable.gitlab.io)

[monty13]

Thanks, I will have a look.

In terms of output, I'd expect to get a DT like the one below.

So, upon a conditional check on the subtracted value, the current row would be tagged as "keep" or NA