Author: Drew Shea, Created: 2026-01-07

Routine Input Parameters

Input Parameters are the configuration values and data sources you provide to a routine when executing a run. They tell the routine what data to process, how to process it, and what options to enable.

---

What Are Input Parameters?

Every routine method requires specific inputs to do its work. Input parameters are how you communicate your intentions to the routine:

• What data should be analyzed

• Which columns serve what purpose

• What settings to use

• What thresholds to apply

Regardless of how you provide them, input parameters follow the same structure and are validated the same way.

---

How Parameters Are Provided

There are two primary ways to provide input parameters to a routine run:

Guided Workflow

The Component Workflow presents parameters through a step-by-step user interface. This approach:

• Groups related parameters into logical steps

• Presents them in a guided sequence

• Validates inputs at each step before proceeding

• Allows earlier choices to influence later options (dynamic dropdowns)

This is the typical experience when working through the web application.

Direct Input

Parameters can also be provided directly as a complete set of values - without stepping through a guided workflow. This approach:

• Supplies all parameters at once

• Skips the step-by-step navigation

• Validates all parameters together before execution

• Is common when automating routine runs or using APIs

Both approaches ultimately produce the same result: a validated set of parameters ready for execution.

---

The Guided Workflow Experience

When using the step-by-step workflow, parameters are organized into steps (also called "states"):

Benefits of the guided approach:

• Easier for new users to understand what's needed

• Dynamic options update based on previous selections

• Validation feedback at each step

• Contextual help and instructions along the way

---

Types of Input Parameters

Regardless of how they're provided, parameters fall into common categories:

Data Sources

Connect your routine to the data it will process.

Parameter Type	Example
Source table	Sales_2024_Q4
File Input	Sales_2024.parquet
Partitioned File Input (Directory containing multiple files of the same schema)	Sales_2024/

Column Mappings

Specify which columns serve specific purposes.

Parameter Type	Example
Date column	"OrderDate"
Target column	"SalesAmount"
Grouping columns	["Region", "ProductCategory"]
Columns to ignore	["InternalID", "Notes"]

Configuration Settings

Control how the routine behaves.

Parameter Type	Example
Number of clusters	5
Algorithm selection	"Isolation Forest"
Normalization	true / false

Thresholds and Boundaries

Set numeric limits for the analysis.

Parameter Type	Example
Anomaly threshold	0.95
Confidence level	0.90
Minimum/maximum values	1 to 20

Feature Toggles

Enable or disable optional capabilities.

Parameter Type	Example
Include visualizations	true
Deterministic mode	true
Generate report	true

---

Parameter Validation

All input parameters are validated before a run begins—regardless of how they're provided.

What Gets Validated

Check	Example
Required fields	"Source data connection is required"
Type correctness	Number fields contain numbers
Range constraints	Value must be between 2 and 20
Length constraints	Text must be 1-100 characters
Pattern matching	Must match expected format
Logical consistency	Selected columns exist in the data source

When Validation Occurs

Approach	Validation Timing
Guided workflow	At each step, and again before submission
Direct input	All at once before execution begins

If validation fails, clear error messages explain what's wrong and how to fix it.

---

Parameter Documentation

Each parameter is documented with metadata that helps you understand what's expected:

Metadata	Purpose
Title	Human-readable name for the parameter
Description	Brief explanation of what it does
Tooltip	Additional guidance and validation rules
Default	Pre-set value if you don't specify one

This documentation is available both in the guided workflow UI and in routine documentation.

---

Default Values

Many parameters have sensible defaults:

• Pre-filled values — Common or recommended settings

• Optional parameters — Can be omitted if not applicable

• Computed defaults — Some defaults depend on your data

Tip: Review defaults before running. They're good starting points but may need adjustment for your specific use case.

---

Working with Parameters Effectively

Know Your Routine

• Read the routine documentation before starting

• Understand what each method expects

• Review example use cases

Know Your Data

• Understand your data structure (columns, types)

• Have data connections configured and accessible

• Know which columns map to which purposes

Iterate and Refine

• Start with defaults when unsure

• Review results and adjust parameters

• Compare runs to understand parameter impact

• Parameters from each run are preserved for reference

---

Summary

Aspect	Key Takeaway
Definition	Configuration values that control routine execution
Provision methods	Guided workflow (step-by-step) or direct input (all at once)
Categories	Data sources, column mappings, settings, thresholds, toggles
Validation	Always validated before execution, regardless of input method
Documentation	Each parameter has title, description, and constraints
Defaults	Sensible defaults provided where possible