CODEWITHSUNDEEP
node.jsrethinkdb
nha
nha

Reputation: 18005

rethinkdb frequency of values

What is the way to implement a "frequency" query in reThinkDB for some value ?

Example data :

[{
 akey : "1",
 anotherkey : "2"
}, {
 akey : "1",
 anotherkey : "2"
}, {
 akey : "AAA",
 anotherkey : "BBB"
}, {
 akey : "CCC",
 anotherkey : "CCC"
}]

Should yield for akey as parameter :

{
  "1"   : 2,
  "AAA" : 1,
  "CCC" : 1
}

Should yield for anotherkey as parameter :

{
  "2"   : 2,
  "BBB" : 1,
  "CCC" : 1
}

Upvotes: 1

Views: 50

Answers (1)

kureikain
kureikain

Reputation: 2314

You can try something like this:

r.expr([{
 akey : "1",
 anotherkey : "2"
}, {
 akey : "1",
 anotherkey : "2"
}, {
 akey : "AAA",
 anotherkey : "BBB"
}, {
 akey : "CCC",
 anotherkey : "CCC"
}]).map(function(doc) {
  return r.branch(doc.keys().contains('akey'),{'value': doc('akey')},{})
})
.group('value')
.count()
.ungroup()
.map(function(doc) {
  return r.object(doc('group'), doc('reduction'))
})
.reduce(function(left, right) {
    return left.merge(right)
})

Another way to do it without reduce is:

r.expr([{
 akey : "1",
 anotherkey : "2"
}, {
 akey : "1",
 anotherkey : "2"
}, {
 akey : "AAA",
 anotherkey : "BBB"
}, {
 akey : "CCC",
 anotherkey : "CCC"
}]).map(function(doc) {
  return r.branch(doc.keys().contains('anotherkey'),{'value': doc('anotherkey')},{})
})
.group('value')
.count()
.ungroup()
.map(function(doc) {
  return [doc('group'), doc('reduction')]
})
.coerceTo('object')

However, I like reduce way because it maps to how I think about the whole process.

Upvotes: 1

Related Questions