---
description: The subtract compute subtracts a number or another column's value from every value of a column.
date_published: 2026-08-13
date_modified: 2026-08-13
canonical_url: https://metricshub.org/community-connectors/develop/computes/subtract.html
---

# subtract (Compute)

On this Page

- [When To Use](#when-to-use)
- [Syntax](#syntax)
- [Properties](#properties)
- [Table Transformation Example](#table-transformation-example)
- [Recommended Pattern](#recommended-pattern)
- [Common Mistakes](#common-mistakes)
- [Community Examples](#community-examples)

## When To Use

Use `subtract` to subtract an operand from every value of one column, row by row. The operand is either a literal number or, with the `$n` syntax, the value of another column of the same row. Typical uses are deriving a “used” quantity from a total and a free counter (used = size − free) or removing an idle component from an aggregate counter. The column values must be numeric.

## Syntax

```yaml
sources:
  fileSystemInformation:
    type: wmi
    namespace: root\CIMv2
    query: SELECT DeviceID,FreeSpace,FreeSpace,Size,Size,VolumeName,FileSystem FROM Win32_LogicalDisk WHERE DriveType = 3
    computes:
    # UsedSpace = Size - FreeSpace
    - type: subtract
      column: 4
      value: $2
```

## Properties

| Property | Required | Default | Description |
| --- | --- | --- | --- |
| `type` | Yes | None | `subtract`. |
| `column` | Yes | None | 1-based index of the column to update. Each value in this column is replaced by `column value - value`. |
| `value` | Yes | None | Operand subtracted from the column value. Either a literal number (e.g. `100`) or a `$n` reference to another column of the same row (e.g. `$2` for the 2nd column). |

## Table Transformation Example

With `column: 4` and `value: $2`, column 4 of each row becomes `column 4 - column 2`:

#### Input

| DeviceID | FreeSpace | FreeSpace | Size | Size | VolumeName | FileSystem |
| --- | --- | --- | --- | --- | --- | --- |
| C: | 100 | 100 | 500 | 500 | System | NTFS |
| D: | 300 | 300 | 400 | 400 | Data | NTFS |

#### Result

| DeviceID | FreeSpace | FreeSpace | UsedSpace | Size | VolumeName | FileSystem |
| --- | --- | --- | --- | --- | --- | --- |
| C: | 100 | 100 | 400 | 500 | System | NTFS |
| D: | 300 | 300 | 100 | 400 | Data | NTFS |

Note how the source query selects `Size` twice: one copy is turned into the used space while the original total remains available for the mapping.

## Recommended Pattern

- Select a field twice in the source query (or use `duplicateColumn`) when you need both the raw value and the subtracted result downstream.
- Keep the operand column intact and reference it with `value: $n`; `subtract` only modifies `column`.
- Comment the resulting column layout above the compute so the mapping indexes stay reviewable.

## Common Mistakes

- Reversing the operands: `subtract` computes `column - value`, never `value - column`. For `constant - column`, first `multiply` the column by `-1`, then `add` the constant.
- Pointing `column` or `$n` at indexes that were shifted by an earlier compute such as `keepColumns` or `awk`.
- Applying `subtract` to a column that contains non-numeric text such as an empty string or a label.

## Community Examples

- [Windows](https://github.com/metricshub/community-connectors/blob/main/src/main/connector/system/Windows/Windows.yaml)
- [Redfish](https://github.com/metricshub/community-connectors/blob/main/src/main/connector/hardware/Redfish/Redfish.yaml)

From `Windows.yaml`, deriving used memory from total and free, included directly from the connector source:

```
# Finding used memory and making a copy for utilization
- type: subtract
  column: 5
  value: $1
```