def test_sample_server():
pop = DistributedPopulation(XgboostIndividual,
size=100,
additional_parameters={'kfold': 3},
maximize=False,
host='localhost',
user='******',
password='******')
ga = GeneticAlgorithm(pop)
ga.run(10)
def test_california_housing_xgb():
data = fetch_california_housing()
y_train = data.target
x_train = data.data
model = XgboostModel(x_train, y_train, kfold=3, num_boost_round=10)
pop = Population(
XgboostIndividual, x_train, y_train, size=5,
additional_parameters={'model': model},
maximize=False
)
ga = GeneticAlgorithm(pop)
ga.run(3)
def test_california_housing_xgb_grid():
data = fetch_california_housing()
y_train = data.target
x_train = data.data
grid = {
'eta': [0.1, 0.2],
'max_depth': range(3, 5),
'colsample_bytree': [0.80, 0.85],
}
model = XgboostModel(x_train, y_train, kfold=3, num_boost_round=10)
pop = GridPopulation(
XgboostIndividual, x_train, y_train, genes_grid=grid,
additional_parameters={'model': model}, maximize=False
)
ga = GeneticAlgorithm(pop)
ga.run(3)
#!/usr/bin/env python
"""
Test the genetic algorithm on a single machine over the
California Housing data using a random population.
"""
import os
import sys
sys.path.insert(0,
os.path.abspath(os.path.join(os.path.dirname(__file__), '..')))
if __name__ == '__main__':
from sklearn.datasets import fetch_california_housing
from gentun import GeneticAlgorithm, Population, XgboostIndividual
data = fetch_california_housing()
y_train = data.target
x_train = data.data
pop = Population(XgboostIndividual,
x_train,
y_train,
size=100,
additional_parameters={'kfold': 3},
maximize=False)
ga = GeneticAlgorithm(pop)
ga.run(10)