I have a collection with fields like "servicereqesttype", "zipcode", "date"
I want to fing the 3 most common "servicerequesttype" per zipcode for a specific day.
db.event.aggregate([
{
$match: {
creationdate: "2011-01-01"
}
},
{
$project: {
zipcode: "$zipcode",
servicerequesttype: "$servicerequesttype"
}
},
{
$group: {
_id: {
zipcode: "$zipcode",
servicerequesttype: "$servicerequesttype"
},
zipcode: {
$first: "$zipcode"
},
servicerequesttype: {
$first: "$servicerequesttype"
},
count: {$sum: 1}
}
},
{
$sort: {
"zipcode": -1,
"count": -1
}
},
{
$project: {
_id: 0,
zipcode: "$zipcode",
servicerequesttype: "$servicerequesttype",
count: "$count"
}
}
])
now all I have to is to select only 3 per zipcode and I need some help, maybe I have to use $bucket or $map...
db.event.aggregate([
{
$match: {
creationdate: "2011-01-01"
}
},
{
$project: {
zipcode: "$zipcode",
servicerequesttype: "$servicerequesttype"
}
},
{
$group: {
_id: {
zipcode: "$zipcode",
servicerequesttype: "$servicerequesttype"
},
zipcode: {
$first: "$zipcode"
},
servicerequesttype: {
$first: "$servicerequesttype"
},
count: {$sum: 1}
}
},
{
$sort: {
"zipcode": -1,
"count": -1
}
},
{
$project: {
_id: 0,
zipcode: "$zipcode",
servicerequesttype: "$servicerequesttype",
count: "$count",
arrayOfTypes: "$array1",
arrayOfIncidents: "$array2"
}
},
{
$group: {
_id: "$zipcode",
arrayOfTypes: {
$push: {type: "$servicerequesttype", count: "$count"}
}
}
},
{
$project: {
_id: "$_id",
array: {
$slice: ["$arrayOfTypes", 3]
}
}
},
{
$sort: {
"_id": -1
}
}
])
Related
I am running the following aggregation pipeline:
const agg = [
{
'$match': {
'aaa': 'bbb'
}
}, {
'$group': {
'_id': '',
'total': {
'$sum': '$num'
}
}
}
];
My problem is, when $match matches nothing, the pipeline returns 0 documents. How do I get the pipeline to always return 1 document?
In MongoDB version 6.0 you can do it like this one:
db.collection.aggregate([
{ $match: { aaa: 'bbb' } },
{
$group: {
_id: null,
total: { $sum: "$num" }
}
},
{
$densify: {
field: "total",
range: { step: 1, bounds: [0, 0] }
}
},
{ $set: { _id: { $cond: [{ $eq: [{ $type: "$_id" }, "missing"] }, MaxKey, "$_id"] } } },
{ $sort: { _id: 1 } },
{ $limit: 1 }
])
In version < 6.0 you can try this one:
db.collection.aggregate([
{
$facet: {
data: [
{ $match: { aaa: 'bbb' } },
{ $group: { _id: null, total: { $sum: "$num" } } }
],
default: [
{ $limit: 1 },
{ $group: { _id: null, total: { $sum: 0 } } },
{ $set: { _id: MaxKey } }
]
}
},
{ $replaceWith: { $mergeObjects: [{ $first: "$default" }, { $first: "$data" }] } },
])
Or this one:
db.collection.aggregate([
{ $match: { aaa: 'bbb' } },
{ $group: { _id: null, total: { $sum: "$num" } } },
{
$unionWith: {
coll: "collection",
pipeline: [
{ $limit: 1 },
{ $set: { _id: MaxKey, total: 0 } },
{ $project: { _id: 1, total: 1 } }
]
}
},
{ $sort: { _id: 1 } },
{ $limit: 1 }
])
code currently in use
db.events.aggregate([
{ '$match':
{ thingId: 'node_0002',
eventState: 'test',
deviceId: {'$in': ['ARCT-1', 'ARCT-2', 'ARSS-1', 'ARSS-2', 'ARSW-1', 'ARCO-1']},
collectTimeText: {'$gte': '2022-04-01T00:00:00+09:00', '$lte': '2022-04-19T23:59:59+09:00' }}},
{ '$group': {
_id: {'$dateFromString': { dateString: { '$substr': [ '$collectTimeText', 0, 10 ] }}
},
list: { '$addToSet': {'deviceId':'$deviceId'} }},
},
{'$unwind':'$list'},
{ '$group': {
_id: {'$dateToString': {'format': "%Y-%m", date: "$_id"}
},
list: { '$push': '$list' }},
},
{'$sort': { _id: 1 } }], {})
result of using the code
{ _id: '2022-04',
list:
[ { deviceId: 'ARCT-2' },
{ deviceId: 'ARSS-1' },
{ deviceId: 'ARCT-1' },
{ deviceId: 'ARSW-1' },
{ deviceId: 'ARCO-1' },
{ deviceId: 'ARSS-2' },
{ deviceId: 'ARCT-2' },
{ deviceId: 'ARSS-1' },
{ deviceId: 'ARCT-1' },
{ deviceId: 'ARSW-1' },
{ deviceId: 'ARCO-1' },
{ deviceId: 'ARSS-2' },
{ deviceId: 'ARCT-2' },
{ deviceId: 'ARSS-1' },
{ deviceId: 'ARCT-1' },
{ deviceId: 'ARSW-1' },
{ deviceId: 'ARCO-1' },
{ deviceId: 'ARSS-2' } ] }
I want an output like the code below. Help
{ _id: '2022-04',
list:
{ 'ARCO-1': 3,
'ARCT-1': 3,
'ARSS-1': 3,
'ARCT-2': 3,
'ARSS-2': 3,
'ARSW-1': 3 } }
This is the code for generating statistics.
How do I get the results I want?
Did I write the code wrong in the first place?
When I added $project , I got the following result. How to check the count
db.events.aggregate([
{ '$match':
{ thingId: 'node_0002',
eventState: 'test',
deviceId: {'$in': ['ARCT-1', 'ARCT-2', 'ARSS-1', 'ARSS-2', 'ARSW-1', 'ARCO-1']},
collectTimeText: {'$gte': '2022-04-01T00:00:00+09:00', '$lte': '2022-04-19T23:59:59+09:00' }}},
{ '$group': {
_id: {'$dateFromString': { dateString: { '$substr': [ '$collectTimeText', 0, 10 ] }}
},
list: { '$addToSet': {'deviceId':'$deviceId'} }},
},
{'$unwind':'$list'},
{ '$group': {
_id: {'$dateToString': {'format': "%Y-%m", date: "$_id"}
},
list: { '$push': '$list' }},
},
{'$project':{
'list':{
'$arrayToObject':{
'$map': {
'input': '$list',
'as': 'el',
'in': {
'k': '$$el.deviceId',
'v': {'$sum':1}
}
}
}
}
}},
{'$sort': { _id: 1 } }], {})
Below is the result of adding $project.
{ _id: '2022-04',
list:
{ 'ARCO-1': 1,
'ARCT-1': 1,
'ARSS-1': 1,
'ARCT-2': 1,
'ARSS-2': 1,
'ARSW-1': 1 } }
I don't know what to do.... It seems like I'm almost there, but I don't know...
Maybe not all...?
$group twice and then $arrayToObject
db.collection.aggregate([
{
$group: {
_id: {
date: "$date",
deviceId: "$deviceId"
},
count: {
$sum: 1
}
}
},
{
$group: {
_id: "$_id.date",
list: {
$push: {
k: "$_id.deviceId",
v: "$count"
}
}
}
},
{
$set: {
list: {
$arrayToObject: "$list"
}
}
}
])
mongoplayground
I am trying to finish up a data aggregation pipeline and having issues getting the data into the correct format. I'm not even sure if this is possible to do in one pipeline.
The original data looks like this:
[
{
answers: {
'question1': 'a',
'question2': 'c',
'question3': ['a','b'],
'question4': 1
},
createdAt: 2022-03-04T07:30:40.517Z,
},
{
answers: {
'question1': 'b',
'question2': 'c',
'question3': ['a','c']
'question4': 2
},
createdAt: 2022-03-04T07:30:40.518Z,
}
]
I've got my pipeline so far with this:
{ $project: {
"answers": { $objectToArray: "$answers" },
"date": { $dateToString: { format: "%Y-%m-%d", date: "$createdAt" }}
}},
{ $unwind: "$answers" },
{ $unwind: "$answers.v" },
{
$group: {
_id: { answers : "$answers", date: "$date"},
c: { $sum: 1 }}
},
and the data now looks like this:
{
_id: {
answers: { k: 'q3', v: 'b' },
date: '2022-03-04'
},
count: 1
},
{
_id: {
answers: { k: 'q3', v: 'a' },
date: '2022-03-04'
},
count: 2
},
{
_id: {
answers: { k: 'q4', v: 1 },
date: '2022-03-04'
},
count: 1
},
{
_id: {
answers: { k: 'q1', v: 'b' },
date: '2022-03-04'
},
count: 1
},
{
_id: {
answers: { k: 'q4', v: 2 },
date: '2022-03-04'
},
count: 1
},
{
_id: {
answers: { k: 'q2', v: 'c' },
date: '2022-03-04'
},
count: 2
},
{
_id: {
answers: { k: 'q3', v: 'c' },
date: '2022-03-04'
},
count: 1
},
{
_id: {
answers: { k: 'q1', v: 'a' },
date: '2022-03-04'
},
count: 1
}
I would like to get a result that looks something like this:
{
'dates': [
{
'date': '2022-03-04',
'q1': { 'a': 1, 'b': 1 }
'q2': { 'c': 2 },
'q3': { 'a': 2, 'b': 1, 'c': 1 },
'q4': { '1': 1, '2': 1 }
}
]
'totals': { // this would be the totals across all the dates
'q1': { 'a': 1, 'b': 1 }
'q2': { 'c': 2 },
'q3': { 'a': 2, 'b': 1, 'c': 1 },
'q4': { '1': 1, '2': 1 }
}
}
any help would be greatly appreciated, even if I can't get both the totals and breakdown in 1 query.
here is the mongoplaygroud I've been working on
Not that simple. An important stage you have to use is $facet in order to get totals and dates
Maybe with $setWindowFields the aggregation pipeline could be a little simpler, but that a quick guess.
db.collection.aggregate([
{
$project: {
_id: 0,
answers: { $objectToArray: "$answers" },
date: { $dateToString: { format: "%Y-%m-%d", date: "$createdAt" } }
}
},
{ $unwind: "$answers" },
{ $unwind: "$answers.v" },
{
$group: {
_id: {
answer: "$answers.v",
question: "$answers.k",
date: "$date"
},
count: { $sum: 1 }
}
},
{
$facet: {
dates: [
{
$group: {
_id: { question: "$_id.question", date: "$_id.date" },
count: {
$push: {
k: { $toString: "$_id.answer" },
v: "$count"
}
}
}
},
{
$group: {
_id: "$_id.date",
count: {
$push: {
k: "$_id.question",
v: { $arrayToObject: "$count" }
}
}
}
},
{
$replaceWith: {
$mergeObjects: [
{ date: "$_id" },
{ $arrayToObject: "$count" }
]
}
}
],
totals: [
{
$group: {
_id: { answer: "$_id.answer", question: "$_id.question" },
v: { $push: "$count" }
}
},
{
$group: {
_id: "$_id.question",
count: {
$push: {
k: { $toString: "$_id.answer" },
v: { $sum: "$v" }
}
}
}
},
{
$project: {
_id: 0,
k: "$_id",
v: { $arrayToObject: "$count" }
}
}
]
}
},
{ $set: { totals: { $arrayToObject: "$totals" } } }
])
Mongo Playground
I have the following resolver:
const result = await UserPassage.aggregate([
{ $sort: { createdAt: -1 } },
{
$group: {
_id: '$level',
level: { $first: '$level' },
passageId: { $first: '$passageId' },
userId: { $first: '$userId' },
type: { $first: '$type' },
category: { $first: '$category' },
score: { $first: '$score' },
completedStage: { $first: '$completedStage' },
userPassageStatsId: {
_id: { $first: '$_id' },
stats: {
readingTime: { $first: '$readingTime' },
qtdVocab: { $first: '$qtdVocab' },
qtdTestDone: { $first: '$qtdTestDone' },
totalQuiz: { $first: '$totalQuiz' },
progress: { $first: '$progress' },
},
},
},
},
{ $sort: { level: 1 } },
]);
await UserPassageStats.populate(result, { path: 'userPassageStatsId' });
The problem is that I need to populate 'userPassageStatsId' and return it but it's not working well returning the following error:
MongoError: The field 'userPassageStatsId' must be an accumulator object
does anyone knows what I am doing wrong?
$group can only contain _id or accumulator objects like $first, $last, $sum etc. In your case your building nested object and that syntax is not allowed - accumulator has to be on a top level. You can try two approaches, either return flat structure from $group and then reshape using $project:
{
$group: {
_id: '$level',
level: { $first: '$level' },
passageId: { $first: '$passageId' },
userId: { $first: '$userId' },
type: { $first: '$type' },
category: { $first: '$category' },
score: { $first: '$score' },
completedStage: { $first: '$completedStage' },
userPassageStatsId_id: { $first: '$_id' },
readingTime: { $first: '$readingTime' },
qtdVocab: { $first: '$qtdVocab' },
qtdTestDone: { $first: '$qtdTestDone' },
totalQuiz: { $first: '$totalQuiz' },
progress: { $first: '$progress' }
}
},
{
$project: {
_id: 1,
level: 1,
...,
userPassageStatsId: {
_id: "$userPassageStatsId_id",
stats: {
readingTime: "$readingTime",
...
}
}
}
}
or use $$ROOT to capture first object for every group and reshape it using $project:
{
$group: {
_id: '$level',
d: { $first: "$$ROOT" }
}
},
{
$project: {
_id: 1,
level: "$d.level",
...,
userPassageStatsId: {
_id: "$d._id",
stats: {
readingTime: "$d.readingTime",
...
}
}
}
}
I am having a bad time trying to do an aggregation in MongoDB.
I need to cross some infos from each user and as a final result I want a list of users (where there is only one object for each user) and for each object there is some lists with distinct information.
1 - The createdAtList array must be ordered from the oldest to the newest date. The sumOfTotal means the current position total summed up with the previous sumOfTotal (Exemplified in the code below), not just the sum of the total's
2 - The categotyList must be ordered like: category1, category2, category3 ...
3 - The desired final result must be ordered like: user1, user2, user3 ...
Basically I need some help to do the following:
//List of docs from my collection:
[
{
_id: "doc1",
user: "user1",
category: "category1",
createdAt: "2018-01-01T00:00:00.000Z"
},
{
_id: "doc2",
user: "user1",
category: "category2",
createdAt: "2017-12-12T00:00:00.000Z",
},
{
_id: "doc3",
user: "user1",
category: "category1",
createdAt: "2017-12-12T00:00:00.000Z",
},
{
_id: "doc4",
user: "user1",
category: "category2",
createdAt: "2018-01-01T00:00:00.000Z"
},
{
_id: "doc5",
user: "user1",
category: "category3",
createdAt: "2017-11-11T00:00:00.000Z"
}
]
//Desired result:
{
user: "user1",
createdAtList: [ //list ordered by createdAt
{
createdAt: "2017-11-11T00:00:00.000Z",
total: 1,
sumOfTotal: 0
}
{
createdAt: "2017-12-12T00:00:00.000Z",
total: 2,
sumOfTotal: 3 //summed up with the previous
}
{
createdAt: "2018-01-01T00:00:00.000Z",
total: 2,
sumOfTotal: 5 //summed up with the previous
}
],
categotyList: [ //list ordered by category
{
category: "category1",
total: 2
},
{
category: "category2",
total: 2
},
{
category: "category3",
total: 1
}
]
},
...
Is possible to do this in the same aggregate?
I do not think it really makes sense to have the createdAtList.sumOfTotal field. I do not think the fields in an array should be dependent upon a particular order of the elements. If you want some field to contain the sum of the createdAtList.total field, I think there should only be one field (outside of the array). That being said, here is the query I came up with to give you the desired results (using "users" as the name of the collection):
db.users.aggregate([
{
$group: {
_id: {
user: "$user",
createdAt: "$createdAt"
},
total: { $sum: 1 },
category: { $push: "$category" }
}
},
{
$project: {
_id: 0,
user: "$_id.user",
createdAt: "$_id.createdAt",
total: "$total",
category: 1
}
},
{ $unwind: "$category" },
{
$group: {
_id: {
user: "$user",
category: "$category"
},
catTotal: { $sum: 1 },
createdAtList: {
$push: {
createdAt: "$createdAt",
total: "$total"
}
}
}
},
{
$project: {
_id: 0,
user: "$_id.user",
createdAtList: 1,
category: "$_id.category",
catTotal: 1
}
},
{ $unwind: "$createdAtList" },
{
$group: {
_id: "$user",
createdAtList: {
$addToSet: "$createdAtList"
},
categoryList: {
$addToSet: {
category: "$category",
total: "$catTotal"
}
}
}
},
{ $unwind: "$createdAtList" },
{ $sort: { "createdAtList.createdAt": 1 } },
{
$group: {
_id: "$_id",
createdAtList: {
$push: "$createdAtList"
},
categoryList: {
$first: "$categoryList"
}
}
},
{ $unwind: "$categoryList" },
{ $sort: { "categoryList.category": 1 } },
{
$group: {
_id: "$_id",
createdAtList: {
$first: "$createdAtList"
},
categoryList: {
$push: "$categoryList"
}
}
},
{
$project: {
_id: 0,
user: "$_id",
createdAtList: 1,
sumOfTotal: { $sum: "$createdAtList.total" },
categoryList: 1
}
},
{ $sort: { user: 1 } },
]).pretty()