MongoDB $setUnion Operator

Example 1 — Sample Data

Imagine two students. Asha learned HTML/CSS then CSS/JS. Ben learned Python then Python/SQL. We save both in a collection named students.

db.students.insertMany([
  {
    _id: 1,
    name: "Asha",
    courseA: ["html", "css"],
    courseB: ["css", "js"]
  },
  {
    _id: 2,
    name: "Ben",
    courseA: ["python"],
    courseB: ["python", "sql"]
  }
]);

How the Data Is Stored in MongoDB

After the insert, MongoDB keeps two documents in students. Run find() to see them:

db.students.find()

[
  {
    _id: 1,
    name: "Asha",
    courseA: ["html", "css"],
    courseB: ["css", "js"]
  },
  {
    _id: 2,
    name: "Ben",
    courseA: ["python"],
    courseB: ["python", "sql"]
  }
]

Read those documents like this:

Asha:
  courseA ∪ courseB
  {html, css} ∪ {css, js}  →  {html, css, js}

Ben:
  courseA ∪ courseB
  {python} ∪ {python, sql} →  {python, sql}

How It Works

1. Insert once. insertMany adds two student documents to students.
2. Two skill arrays. Each course contributes its own list; overlap (like css) is expected.
3. find() shows both rows. What you inserted is what you get back—ready for the union next.

Example 2 — Merge Skills

Goal: add a skills field—the unique set of all course skills.

db.students.aggregate([
  {
    $addFields: {
      skills: {
        $setUnion: ["$courseA", "$courseB"]
      }
    }
  }
]);

[
  {
    _id: 1,
    name: "Asha",
    courseA: ["html", "css"],
    courseB: ["css", "js"],
    skills: ["html", "css", "js"]
  },
  {
    _id: 2,
    name: "Ben",
    courseA: ["python"],
    courseB: ["python", "sql"],
    skills: ["python", "sql"]
  }
]

How It Works

1. Read both arrays. Still using courseA and courseB from Example 1.
2. Pass them to $setUnion. Every unique value from either list is kept.
3. Duplicates drop. Shared values like css or python appear once.
4. Keep all fields. $addFields adds skills next to the original arrays (order may vary).

Example 3 — Filter with $expr

Keep only students whose combined skills include js.

db.students.aggregate([
  {
    $match: {
      $expr: {
        $in: [
          "js",
          { $setUnion: ["$courseA", "$courseB"] }
        ]
      }
    }
  },
  {
    $project: {
      name: 1,
      courseA: 1,
      courseB: 1
    }
  }
]);

{
  _id: 1,
  name: "Asha",
  courseA: ["html", "css"],
  courseB: ["css", "js"]
}

How It Works

1. Union first. Build the unique skill set from both courses (same idea as Example 2).
2. Test membership. $in checks whether "js" is in that set.
3. Wrap with $expr. Aggregation operators need $expr inside $match.
4. Keep matches only. Asha stays; Ben is dropped.