Statistical Functions

8 min

The Analytics module makes use of built-in statistical functions to process the input data. Find a processor by the task you want to perform, then follow the link for its parameters.

Transform a value

Processor	What it does
Change	Sets, replaces, deletes, or moves a field's value.
Conversion	Converts common factory units, such as temperature, angle, or timestamp format.
Base Conversion	Converts a number between decimal, binary, octal, and hexadecimal.
Normalization	Scales incoming values into a 0-to-1 range.
Rounding	Rounds a number to a set number of decimal digits.

Filter and smooth a signal

Processor	What it does
Gaussian Filter	Smooths a signal using a Gaussian weighting function while preserving its trend.
Moving Window	Calculates a continuously updating average, maximum, or minimum over a sliding window.
Simple Window	Outputs a raw moving window of recent values, oldest to newest.
Signal Decomposition	Separates a time series into trend, seasonality, and residual components.

Detect patterns and monitor quality

Processor	What it does
Anomaly Detection	Flags data points outside a normal statistical range using the three-sigma rule.
Rise and Fall	Detects when a value crosses a threshold and reports the crossing's duration.
SPC Charts	Assesses process performance and capability against specification limits, flagging Nelson Rule violations.
XMR Chart	Tracks a process against control limits calculated from a stabilized baseline.
Statistical Functions	Tests whether a window of data matches a normal distribution using goodness-of-fit tests.

Predict future values

Processor	What it does
ARIMA Filter	Forecasts the next value using an ensemble of ARIMA models fit to a rolling window.
Linear Prediction	Forecasts future values by fitting a linear trend line to recent data.
Statistical Prediction	Forecasts future values with cyclical patterns using a Fast Fourier Transform.

Route or combine messages

Processor	What it does
Switch	Forwards a message only when it passes a rule you define.
Delay/Rate-Limit	Delays each message or limits how many pass through per time window.
Inputs	Calculates an average, maximum, minimum, or sum across multiple inputs.
Combination Processor	Combines fields from multiple input processors into a single JSON string.
Join Processor	Combines the output of multiple input processors into a single message.

Apply AI and ML models

Processor	What it does
AI Processor	Sends data to an AI model and returns its response, for summarizing, classifying, or extracting structured data.
AI Responses	Sends messages to an AI model and keeps conversation history across calls. Add tools, files, and memory to make it an agent.
AI Vision Processor	Sends an image to an AI model and returns its response.
Feature Extractor	Calculates statistical features, such as average, standard deviation, and skewness, from a window of values.
TensorFlow Processor	Feeds time-series data to a trained TensorFlow model.
TensorFlow Images Processor	Feeds images to a trained TensorFlow model.

Write custom logic

Processor	What it does
Expression	Evaluates a custom mathematical formula against input values.
JavaScript Processor	Adds, removes, or modifies message fields using JavaScript.
JSONata Processor	Reshapes, filters, or combines JSON fields using a JSONata expression.
Tengo Script	Adds, removes, or modifies message fields using a Tengo script.

Read files

Processor	What it does
File Reading Processor	Reads a file's contents and returns them in the output payload.

Docs powered by Archbee