Why do i miss data when i use $project

Working with Data
1

hi,
does anyone have an idea, what’s the reason for missing data in my aggregation?
This is my pipeline:

[
  {
    $lookup: {
      as: "resultcheck",
      foreignField: "checkId",
      from: "checks",
      localField: "checkId",
    },
  },
  {
    $lookup: {
      as: "resultDepartment",
      foreignField: "DepId",
      from: "department",
      localField: "DepartmentId",
    },
  },
  {
    $unwind: {
      path: "$X",
    },
  },
  {
    $unwind: {
      path: "$X.X",
    },
  },
  {
    $unwind: {
      path: "$resultDepartment",
    },
  },
  {
    $unwind: {
      path: "$resultcheck",
    },
  },
  {
    $match: {
      $and: [
        {
          "X.X.TotalType":
            "Forecast",
          "X.X.FiscalYear":
            {
              $ne: null,
            },
        },
      ],
    },
  },
  {
    $project: {
      _id: true,
      number: "$Number",
      responsible: "$rResponsible",
      calendarMonth: "$Period",
      department: "$resultDepartment.Name",
      finishDate: "$FinishDate",
      xYear:   "$X.X.Year",
      year: "$Year",
      level: "$Level",
      check: "$resultcheck.check",
      period: {
        $switch: {
          branches: [
            {
              case: {
                $eq: ["$Period", 10],
              },
              then: 1,
            },
            {
              case: {
                $eq: ["$Period", 11],
              },
              then: 2,
            },
            {
              case: {
                $eq: ["$Period", 12],
              },
              then: 3,
            },
            {
              case: {
                $eq: ["$Period", 1],
              },
              then: 4,
            },
            {
              case: {
                $eq: ["$Period", 2],
              },
              then: 5,
            },
            {
              case: {
                $eq: ["$Period", 3],
              },
              then: 6,
            },
            {
              case: {
                $eq: ["$Period", 4],
              },
              then: 7,
            },
            {
              case: {
                $eq: ["$Period", 5],
              },
              then: 8,
            },
            {
              case: {
                $eq: ["$Period", 6],
              },
              then: 9,
            },
            {
              case: {
                $eq: ["$Period", 7],
              },
              then: 10,
            },
            {
              case: {
                $eq: ["$Period", 8],
              },
              then: 11,
            },
            {
              case: {
                $eq: ["$Period", 9],
              },
              then: 12,
            },
          ],
        },
      },
      title: "$Title",
      type: "$X.X.TotalType",
      value:
        "$X.X.Value",
    },
  },
]

i’ve got from over 17k documents only 16k.
The original document contains a lot of objects with arrays inside the objects, which i resolve with unwind i.e.

Can someone help?

Kind Regards
LJ

PS: i have anonymized some parts of this code

2

Project should not limit documents output, have you commenting out all stages and gradually running and counting document with each stage gradually introduced?

3

almost done. If all is commented out i’ve got all docs.
I also checked more than one time the lookups, to be sure that nothin is happen there. Could not find any issues after lookup or match

4

Your switch statement could also be simplified using the modulus operator:

db.getCollection("numbers").aggregate([
{
    $addFields:{
        newNum:{
            $add:[
                {
                    $mod:[
                        {$add:['$num', 2]},
                        12
                    ]
                },
                1
            ]
        }            
    }
}
])
5

Check the following $unwind option:

1 Like
6

It seems to be an issue with the X.X Values which where needed.

i changed the $unwind to:

{
$unwind: {
path: "myPath",
includeArrayIndex: "string",
preserveNullAndEmptyArrays: true,
},
}

, changed the match statement to 2 single statements and the missing data popped up and where available now…

Thy very much :slight_smile:

1 Like
7

i will try, thy for the tip