def test_multiple_prediction():
test_data = load_dataset(file_name='test.csv')
multiple_test_input = test_data.to_json(orient='records')
subject = predict.make_prediction(multiple_test_input)
assert subject is not None
assert len(subject) == 482
def test_make_single_prediction():
test_data = load_dataset(file_name='test.csv')
single_test_input = test_data[0:1].to_json(orient='records')
subject = predict.make_prediction(single_test_input)
assert subject is not None
assert isinstance(subject[0], float)
def run_training():
"""Train the model."""
# read training data
data = data_management.load_dataset(
file_name=configuracion.TRAINING_DATA_FILE)
y = data[configuracion.TARGET]
model = pipeline.preprocessor_pipe.fit(data, y)
data_management.save_pipeline(model_to_persist=model)
_logger.debug(f'Training model version: {configuracion._version} ')
from azureml.core import Workspace
from azureml.core import Experiment
from regression_model.data_management import load_dataset
from regression_model.config import configuracion
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score
ws = Workspace.from_config()
experiment = Experiment(workspace=ws, name="finalexp2")
data = load_dataset(file_name=configuracion.TRAINING_DATA_FILE)
y = data[configuracion.TARGET]
X_train, X_test, y_train, y_test = train_test_split(data, y, test_size=0.2, random_state=66)
#from data_management import load_pipeline
#import configuracion
from sklearn.externals import joblib
from sklearn.pipeline import Pipeline
from sklearn.preprocessing import StandardScaler
from sklearn.ensemble import RandomForestClassifier
from sklearn.base import BaseEstimator, TransformerMixin
class ImputeNa(BaseEstimator, TransformerMixin):
""" Replace nan values with 'missing' """
def __init__(self, variables=None) -> None:
if not isinstance(variables, list):
self.variables = [variables]