jq summarize similar elements - json

Question

I want to use jq to summarize elements with the same description field value.
So that each unique field value of description gets it's own element with the amount field summarized.
I'm using jq 1.5
Before:
{ "frames":
[
{ "description": "Stuff", "amount": 8 },
{ "description": "Stuff", "amount": 4 },
{ "description": "other_stuff", "amount": 2 },
{ "description": "more_stuff", "amount": 20 }
]
}
After:
{ "frames":
[
{ "description": "Stuff": 12 },
{ "description": "other_stuff", "amount": 2 },
{ "description": "more_stuff", "amount": 20 }
]
}

Answer 1

With your input, the following program produces valid JSON in the spirit of the question:
def sum_by(f;g): reduce .[] as $x ({}; .[$x|f] += ($x|g));
.frames |= sum_by(.description; .amount)
Using sum_by is more efficient than using group_by for various reasons.
Output
{
"frames": {
"Stuff": 12,
"other_stuff": 2,
"more_stuff": 20
}
}
Variant
If you want description and amount in the output, you could tweak the above as follows:
.frames |=
(sum_by(.description; .amount)
| to_entries
| map( {description: .key, amount: .value} ))
The output would then be:
{
"frames": [
{
"description": "Stuff",
"amount": 12
},
{
"description": "other_stuff",
"amount": 2
},
{
"description": "more_stuff",
"amount": 20
}
]
}