Numerical Filter

---

Description

The Numerical Filter processor filters events based on numerical comparisons with a specified threshold. It supports various comparison operations and is ideal for:

• Threshold-based event filtering
• Range-based data selection
• Outlier detection
• Value-based event routing

---

Required Input

A data stream containing at least one numerical field to filter on.

---

Configuration

Field

• Select the numerical field to apply the filter operation on
• The field must contain numeric values

Operation

Choose from the following comparison operators:

• < (Less than)
• <= (Less than or equal)
• > (Greater than)
• >= (Greater than or equal)
• == (Equal)
• != (Not equal)

Threshold Value

• Specify the numerical threshold to compare against
• The value must be a valid number

Output

The processor forwards the input event only if the numerical comparison evaluates to true.

Example

Input Events

{
  "temperature": 25.5,
  "timestamp": 1586380104915
}
{
  "temperature": 26.0,
  "timestamp": 1586380105015
}
{
  "temperature": 25.8,
  "timestamp": 1586380105115
}

Example 1: Greater Than Filter

Configuration:

• Field: temperature
• Operation: >
• Threshold Value: 25.8

Output Events:

{
  "temperature": 26.0,
  "timestamp": 1586380105015
}

Example 2: Range Filter

Configuration:

• Field: temperature
• Operation: >=
• Threshold Value: 25.5

Output Events:

{
  "temperature": 25.5,
  "timestamp": 1586380104915
}
{
  "temperature": 26.0,
  "timestamp": 1586380105015
}
{
  "temperature": 25.8,
  "timestamp": 1586380105115
}

Example 3: Exact Match Filter

Configuration:

• Field: temperature
• Operation: ==
• Threshold Value: 25.8

Output Events:

{
  "temperature": 25.8,
  "timestamp": 1586380105115
}

Use Cases

1. Threshold Monitoring

   • Alert on values exceeding limits
   • Filter out normal readings
   • Monitor critical thresholds
   • Track value ranges

2. Data Quality

   • Remove outliers
   • Filter invalid measurements
   • Ensure value ranges
   • Validate sensor data

Notes

• The processor performs exact numerical comparisons
• For equality checks, a small epsilon (0.000001) is used to handle floating-point precision
• Events that don't match the filter condition are dropped
• The original event structure is preserved in the output
• All numerical types (integer, float, double) are supported