CODEWITHSUNDEEP
mongodbmongodb-querypymongo
idailylife
idailylife

Reputation: 184

MongoDB query by compound indexed values

Given a collection of rows like

------------------------
date    | time | value
------------------------
20190101| 0930 | 0.00
20190101| 0931 | 1.23
20190102| 0930 | 9.83
20190103| 0745 | NaN
....
------------------------

an I created a compound index (e.g. (date, time)) on the collection. What is the fastest way to query for records whose (date, time) are in the list -- for example,

query items that match:
{date:20190101, time:0931} or {date:20190102, time:0930}
should return
------------------------
date    | time | value
------------------------
20190101| 0931 | 1.23
20190102| 0930 | 9.83
------------------------

There is a $in aggregation but idk how to use it on multiple fields. By the way I am using pymongo.

Upvotes: 0

Views: 51

Answers (1)

Ali Husseinat
Ali Husseinat

Reputation: 116

Let's get an example of a query using $in:

db.collection.find({
    date: {
        $in: [date0, date1, date2]
    }
})

This same operation has the same result as a sequence of $or :

db.collection.find({
    $or: [{
        date: date0
    },
    {
        date: date1
    },
    {
        date: date2
    }]
})

With this in mind, is possible to have the same behavior as an $in statement with multiple fields by translating it to this $or notation:

db.collection.find({
    $or: [{
        date: date0,
        timestamp: time0
    },
    {
        date: date1,
        timestamp: time1
    },
    {
        date: date2,
        timestamp: time2
    }]
})

Upvotes: 1

Related Questions