Sort, Unique & Group with jq (sort_by, unique, group_by)

Example input

[
  { "name": "Ada",   "dept": "eng",   "age": 36 },
  { "name": "Linus", "dept": "eng",   "age": 54 },
  { "name": "Mae",   "dept": "sci",   "age": 28 }
]

Sort by a field

jq 'sort_by(.age)' data.json          # ascending
jq 'sort_by(.age) | reverse' data.json # descending

Sort by multiple keys (array of criteria):

jq 'sort_by(.dept, .age)' data.json

Unique / dedupe

unique sorts and removes duplicates of scalar arrays:

echo '[3,1,2,1,3]' | jq 'unique'     # [1,2,3]

Dedupe objects by a field with unique_by:

jq 'unique_by(.dept)' data.json     # one object per dept

Group by a key

jq 'group_by(.dept)' data.json

Output: an array of arrays, one per dept.

Count per group

jq 'group_by(.dept) | map({ dept: .[0].dept, count: length })' data.json

Output:

[
  { "dept": "eng", "count": 2 },
  { "dept": "sci", "count": 1 }
]

Min / max by a field

jq 'max_by(.age)' data.json     # the oldest record
jq 'min_by(.age)' data.json     # the youngest record

group_by sorts first, so groups come out ordered by the key. To turn groups into a {key: items} object, follow with map({ (.[0].dept): . }) | add. For top-N, sort_by(.age) | reverse | .[:3].