MongoDB aggregation $divide computed fields

Question

I am trying to compute a percentage in a MongoDB query based on computed fields - not sure if this is possible or not. What I'd like to be able to do is calculate the failure percentage: (failed count / total) * 100

Here are a few sample documents:

    {
            "_id" : ObjectId("52dda5afe4b0a491abb5407f"),
            "type" : "build",
            "time" : ISODate("2014-01-20T22:39:43.880Z"),
            "data" : {
                    "buildNumber" : 30,
                    "buildResult" : "SUCCESS"
            }
    },
    {
            "_id" : ObjectId("52dd9fede4b0a491abb5407a"),
            "type" : "build",
            "time" : ISODate("2014-01-20T22:15:07.901Z"),
            "data" : {
                    "buildNumber" : 4,
                    "buildResult" : "FAILURE"
            }
    },
    {
            "_id" : ObjectId("52dda153e4b0a491abb5407b"),
            "type" : "build",
            "time" : ISODate("2014-01-20T22:21:07.790Z"),
            "data" : {
                    "buildNumber" : 118,
                    "buildResult" : "SUCCESS"
            }
    }

Here is the query I am trying to work with. The issue is in the FailPercent/$divide line:

db.col.aggregate([    
    { $match: { "data.buildResult" : { $ne : null } } },
    { $group: {          
        _id: {              
            month: { $month: "$time" },             
            day: { $dayOfMonth: "$time" },             
            year: { $year: "$time" },                       
        },         
        Aborted: { $sum: { $cond :  [{ $eq : ["$data.buildResult", "ABORTED"]}, 1, 0]} },
        Failure: { $sum: { $cond :  [{ $eq : ["$data.buildResult", "FAILURE"]}, 1, 0]} },
        Unstable: { $sum: { $cond : [{ $eq : ["$data.buildResult", "UNSTABLE"]}, 1, 0]} },
        Success: { $sum: { $cond :  [{ $eq : ["$data.buildResult", "SUCCESS"]}, 1, 0]} },
        Total: { $sum: 1 },
        FailPercent: { $divide: [ "Failure", "Total" ] }
    } },     
    { $sort: { "_id.year": 1, "_id.month": 1, "_id.day": 1 } } 
])

Answer 1

You almost got it. Only change that would be required is that you'll have to compute the FailPercent in an additional project phase, because the total is only available after the completion of the group phase. Try this:

db.foo.aggregate([    
    { $match: { "data.buildResult" : { $ne : null } } },
    { $group: {          
        _id: {              
            month: { $month: "$time" },             
            day: { $dayOfMonth: "$time" },             
            year: { $year: "$time" },                       
        },         
        Aborted: { $sum: { $cond :  [{ $eq : ["$data.buildResult", "ABORTED"]}, 1, 0]} },
        Failure: { $sum: { $cond :  [{ $eq : ["$data.buildResult", "FAILURE"]}, 1, 0]} },
        Unstable: { $sum: { $cond : [{ $eq : ["$data.buildResult", "UNSTABLE"]}, 1, 0]} },
        Success: { $sum: { $cond :  [{ $eq : ["$data.buildResult", "SUCCESS"]}, 1, 0]} },
        Total: { $sum: 1 }
    } }, 
    {$project:{Aborted:1, Failure:1, Unstable:1, Success:1, Total:1, FailPercent: { $divide: [ "$Failure", "$Total" ]}}},
    { $sort: { "_id.year": 1, "_id.month": 1, "_id.day": 1 } } 
])