Run Experiments Method (POST)#

This is the core endpoint of the GADEMO API, responsible for executing the genetic algorithm based on custom user parameters and returning structured evolutionary data.

Overview#

Method: POST
Endpoint: /run-experiments
Response Format: JSON

This method takes in a mathematical function and a number of experiments to run. It returns a full record of the algorithm’s behavior across all experiments and generations.

Parameters (Query)#

func_str (string, required): The objective function to be optimized. Must be expressed in valid Python-like math syntax (e.g., sin(x) + cos(y)).
num_experiments (integer, required): The number of independent experiments (runs) to be executed using the same configuration.

Request Body (JSON)#

The body must be a JSON object specifying the algorithm’s configuration:

{
  "num_generations": 3,
  "population_size": 10,
  "crossover_rate": 65,
  "mutation_rate": 0.80,
  "maximize": true,
  "interval": [-100, 100],
  "crossover_type": {
    "one_point": true,
    "two_point": false,
    "uniform": false
  },
  "normalize_linear": false,
  "normalize_min": 0,
  "normalize_max": 100,
  "elitism": false,
  "steady_state": false,
  "steady_state_without_duplicateds": false,
  "gap": 0
}

Endpoint#

Request URL (example):

https://servicosmaxwell.vrac.puc-rio.br/run-experiments?func_str=0.5%20-%20%28sin%28sqrt%28x%5E2%20%2B%20y%5E2%29%29%5E2%20-%200.5%29%20%2F%20%281%20%2B%200.001%20%2A%20%28x%5E2%20%2B%20y%5E2%29%29%5E2&num_experiments=2

Response#

A successful response will return the following fields:

best_experiment_values: Best fitness value obtained in each experiment.
best_individuals_per_generation: Best individual (solution) at each generation.
mean_best_individuals_per_generation: Mean fitness of best individuals across experiments per generation.
best_values_per_generation: Evolution of best fitness values for all experiments.
last_generation_values: Fitness distribution of all individuals in the final generation.
execution_time_seconds: Total time spent running the experiments.

Response Example (using F6 function):

{
  "best_experiment_values": [
    0.5188708575694038,
    0.5123384091156766
  ],
  "best_individuals_per_generation": [
    [
      [-41.96, 73.63],
      [-41.96, 73.63],
      [-43.10, 45.96]
    ],
    [
      [35.77, 36.19],
      [35.77, 36.19],
      [35.77, 36.19]
    ]
  ],
  "mean_best_individuals_per_generation": [
    0.50986,
    0.50986,
    0.51560
  ],
  "best_values_per_generation": [
    [0.5073, 0.5073, 0.5188],
    [0.5123, 0.5123, 0.5123]
  ],
  "last_generation_values": [
    [0.50, 0.51, 0.50, 0.50, 0.50, 0.50, 0.51, 0.50, 0.50, 0.50],
    [0.50, 0.51, 0.50, 0.51, 0.51, 0.51, 0.51, 0.51, 0.51, 0.51]
  ],
  "execution_time_seconds": 0.2145
}

This endpoint powers the entire GADEMO experimentation process and returns all data needed for statistical analysis, visualizations, and result inspection.