# Overview

The Grouped FFT Computation Processor computes Fast Fourier Transformations for datasets grouped by a given time-window size. It uses a divide and conquer algorithm that efficiently decomposes digital signals into frequencies.

# Input

As input, the processor requires a dataset with a timestamp column (of type datetime) and numeric column to which compute the FFT.

# Configuration

*For more information, check the linked section.*

# Output

The processor returns a table with five columns: The sensor column name, the window number and the corresponding FFT properties (__frequency, power magnitude and phase__).

# Example

## Example Input

## Workflow

In the following workflow, the Ordering processor is used to order the output data by ascending window number. This doesn't affect the result, it just makes it easier to interpret.

## Example Configuration

## Result

# Additional Information

This section clarifies some of the concepts appearing in the processor configuration:

- The Hann function of length L is a window function given by
**:**

**Zero padding simply refers to adding zeros****to a time-domain signal to increase its length.****Interpolation is the**process of estimating and inserting missing values in time series data.