How to calculate delta array for a cumulated array in a performant way using JSONata? - jsonata

Question

I am having a json structure consisting of an array of 253 location objects. Each location object has an array reporting the cumulated total over time (array having 67 elements).
I would like to extend each location object in the json structure with an array (delta array) that is reporting the increase based on the cumulated total array.
The following json query is correctly doing this task:
(
covid ~> | locations | {
"new_total" : $map(cumulated_total, function($v, $i, $a) {$a[$i]-($i=0 ? 0 : $a[$i-1])})
} |;
)
The problem with that query is that it takes about 66 seconds to execute on my intel NUC device.
The JSONata Exerciser link containing an example of the json input structure together with the query I have used:
https://try.jsonata.org/BGqIUkWe7
Note that when I open that link it is reporting Expression evaluation timeout: Check for infinite loop because the query also takes too long to execute in my JSONata Exerciser window.

Answer 1

Not sure if it makes much difference, but you could try pulling the function definition up to the top level so that it only gets done once (since you're not relying on any lexical context in the function body):
(
$fn := function($v, $i, $a) {$v-($i=0 ? 0 : $a[$i-1])};
covid ~> | locations | {
"new_total" : $map(cumulated_total, $fn)
} |;
)

Answer 2

Following query is more performant (47 sec total time). So this is 19 sec more performant compared to query in the original post (66 sec total time).
(
covid ~> | locations | {
"new_total" : $map(cumulated_total, function($v, $i, $a) {$v-($i=0 ? 0 : $a[$i-1])})
} |;
)
As you can see I have only changed $a[$i] by $v in the function code.