Navigation

$unsetField (aggregation)

On this page

Definition

$unsetField

New in version 5.0.

Removes a specified field in a document.

You can use $unsetField to remove fields with names that contain periods (.) or that start with dollar signs ($).

$unsetField is an alias for $setField using $$REMOVE to remove fields.

Syntax

$unsetField has the following syntax:

copy 
{
   $unsetField: {
      field: <String>,
      input: <Object>,
   }
}

You must provide the following fields:

Field Type Description
field String Field in the input object that you want to add, update, or remove. field can be any valid expression that resolves to a string constant.
input Object Document that contains the field that you want to add or update. input must resolve to an object, missing, null, or undefined.

Behavior

  • If input evaluates to missing, undefined, or null, $unsetField returns null and does not update input.
  • If input evaluates to anything other than an object, missing, undefined, or null, $unsetField returns an error.
  • If field resolves to anything other than a string constant, $unsetField returns an error.
  • If field doesn’t exist in input, $unsetField adds it.
  • $unsetField doesn’t implicitly traverse objects or arrays. For example, $unsetField evaluates a field value of "a.b.c" as a top-level field "a.b.c" instead of as a nested field, { "a": { "b": { "c": } } }.

Examples

Remove Fields that Contain Periods (.)

Consider the inventory collection:

copy 
db.inventory.insertMany( [
   { _id: 1, item: "sweatshirt", qty: 300, "price.usd": 45.99 },
   { _id: 2, item: "winter coat", qty: 200, "price.usd": 499.99 },
   { _id: 3, item: "sun dress", qty: 250, "price.usd": 199.99 },
   { _id: 4, item: "leather boots", qty: 300, "price.usd": 249.99 },
   { _id: 5, item: "bow tie", qty: 180, "price.usd": 9.99 }
 ] )

Use the $replaceWith pipeline stage and the $unsetField operator to remove the "price.usd" field from each document:

copy 
db.inventory.aggregate( [
   { $replaceWith: {
        $unsetField: {
           field: "price.usd",
           input: "$$ROOT"
   } } }
 ] )

The operation returns the following results:

[
  { _id: 1, item: 'sweatshirt', qty: 300 },
  { _id: 2, item: 'winter coat', qty: 200 },
  { _id: 3, item: 'sun dress', qty: 250 },
  { _id: 4, item: 'leather boots', qty: 300 },
  { _id: 5, item: 'bow tie', qty: 180 }
]

Remove Fields that Start with a Dollar Sign ($)

Consider the inventory collection:

copy 
db.inventory.insertMany( [
   { _id: 1, item: "sweatshirt", qty: 300, "$price": 45.99 },
   { _id: 2, item: "winter coat", qty: 200, "$price": 499.99 },
   { _id: 3, item: "sun dress", qty: 250, "$price": 199.99 },
   { _id: 4, item: "leather boots", qty: 300, "$price": 249.99 },
   { _id: 5, item: "bow tie", qty: 180, "$price": 9.99 }
 ] )

Use the $replaceWith pipeline stage with the $unsetField and $literal operators to remove the "$price" field from each document:

copy 
db.inventory.aggregate( [
   { $replaceWith: {
        $unsetField: {
           field: { $literal: "$price" },
           input: "$$ROOT"
   } } }
] )

The operation returns the following results:

[
  { _id: 1, item: 'sweatshirt', qty: 300 },
  { _id: 2, item: 'winter coat', qty: 200 },
  { _id: 3, item: 'sun dress', qty: 250 },
  { _id: 4, item: 'leather boots', qty: 300 },
  { _id: 5, item: 'bow tie', qty: 180 }
]

Remove A Subfield

Consider the inventory collection:

copy 
db.inventory.insertMany( [
   { _id: 1, item: "sweatshirt", qty: 300, "price": {"usd":45.99, "euro": 38.77 } },
   { _id: 2, item: "winter coat", qty: 200, "price": { "usd": 499.99, "euro": 420.51 } },
   { _id: 3, item: "sun dress", qty: 250, "price": { "usd": 199.99, "euro": 167.70 } },
   { _id: 4, item: "leather boots", qty: 300, "price": { "usd": 249.99, "euro": 210.68 } },
   { _id: 5, item: "bow tie", qty: 180, "price": { "usd": 9.99, "euro": 8.42 } }
 ] )

The "price" field contains a document with two subfields, "usd" and "euro". You cannot use "price.euro" to identify and remove "euro" because MongoDB parses "price.euro" as a top level field name that happens to contain a period (.).

Use the $replaceWith pipeline stage with $setField and a nested $unsetField operation to remove the "euro" field:

copy 
db.inventory.aggregate( [
   { $replaceWith: {
        $setField: {
           field: "price",
           input: "$$ROOT",
           value: {
              $unsetField: {
                 field: "euro",
                 input: { $getField: "price" }
   } } } } }
] )

The operation returns the following results:

[
   { _id: 1, item: "sweatshirt", qty: 300, price: { usd: 45.99 } },
   { _id: 2, item: "winter coat", qty: 200, price: { usd: 499.99 } },
   { _id: 3, item: "sun dress", qty: 250, price: { usd: 199.99 } },
   { _id: 4, item: "leather boots", qty: 300, price: { usd: 249.99 } },
   { _id: 5, item: "bow tie", qty: 180, price: { usd: 9.99 } }
]

See also

$setField

© MongoDB, Inc 2008-present. MongoDB, Mongo, and the leaf logo are registered trademarks of MongoDB, Inc.