Trial Endpoints

Trials in globalMOO represent individual optimization runs. Each trial contains input values, objectives, and results of an optimization attempt.

Available Endpoints

• Read Trial - Retrieve information about a specific trial

Trial Object

A trial object contains the following fields:

{
    "id": 123,
    "project_id": 456,
    "number": 1,
    "case_count": 3,
    "output_count": 3,
    "outputs": [
        [1.2, 3.4, 5.6],
        [7.8, 9.0, 1.2],
        [3.4, 5.6, 7.8]
    ],
    "created_at": "2025-01-31T10:00:00.000Z",
    "updated_at": "2025-01-31T10:00:00.000Z"
}

Field	Type	Description
id	integer	Unique identifier for the trial
project_id	integer	ID of the parent project
number	integer	Trial sequence number (auto-incrementing)
case_count	integer	Number of test cases in trial
output_count	integer	Number of outputs per case
output_cases	array[array[float]]	Array of test output case arrays
objectives	array[Objective]	Array of optimization objectives
created_at	string	ISO 8601 timestamp with millisecond precision
updated_at	string	ISO 8601 timestamp with millisecond precision
disabled_at	string	Optional ISO 8601 timestamp when trial was disabled

Trial Lifecycle

1. A trial is created when output cases are loaded

2. Trial progression:

   • Output cases are collected and validated

   • Objectives are defined with constraints

   • Inverse optimization steps are executed

   • Results are tracked through objectives

3. Trial completion can be determined by:

   • Checking if any objectives are satisfied

   • Examining inverse optimization stop reasons

   • Monitoring L1 norm convergence

Objective Integration

Trials track optimization through Objective objects which contain:

• Target values and types

• Bounds and constraints

• Inverse optimization steps

• Result satisfaction details

Common Use Cases

1. Monitoring output collection progress

2. Retrieving trial output data

3. Analyzing output patterns

4. Managing multiple output cases