def test_bad_yaml_raises_valueerror(self):
self.assertRaises(ValueError, lambda: PipelineGenerator.from_yaml({}))
from doepipeline.designer import BaseExperimentDesigner
import pandas as pd
import numpy as np
class ExampleDesigner(BaseExperimentDesigner):
def __init__(self, *args, **kwargs):
super(ExampleDesigner, self).__init__(*args, **kwargs)
self.design = None
np.random.seed(123456789)
def update_factors_from_response(self, response):
self.design += 1
return self.design
def new_design(self, factor_settings=None):
data = np.random.randint(10, size=(2, len(self.factors)))
self.design = pd.DataFrame(data,
index=['A', 'B'],
columns=self.factors.keys())
return self.design
if __name__ == '__main__':
generator = PipelineGenerator.from_yaml('example_pipeline.yaml')
designer = generator.new_designer_from_config(ExampleDesigner)
design = designer.new_design()
pipeline = generator.new_pipeline_collection(design)
executor = LocalPipelineExecutor()
results = executor.run_pipeline_collection(pipeline)
design = designer.update_factors_from_response(results)
def test_creating_from_yaml_doesnt_crash(self):
yaml_generator = PipelineGenerator.from_yaml(self.yaml_path)
with open(self.yaml_path) as f:
buffer_generator = PipelineGenerator.from_yaml(f)