prediction.extend(
rapidminer_train(go_url, gouser, gopassword, input_train_data, label,
cost_matrix, high_value, low_value,
selection_criteria, max_min_crietria_selector,
platform))
prediction.extend(
rapidminer_score(go_url, gouser, gopassword, input_score_data, label,
depID))
print('returning result')
return prediction
print('Deploying Quick Model')
tabclient.deploy('Rapidminer_Quick_Training',
rapidminer_quick_training,
'Quickly Trains a model for predictions',
override=True)
print('Deploying Training and Score')
tabclient.deploy('RapidMiner_Train_And_Score',
rapidminer_train_and_score,
'Trains and Returns a dataset with predictions',
override=True)
print('Deploying Score')
tabclient.deploy('RapidMiner_Score',
rapidminer_score,
'Returns a dataset with predictions',
override=True)
print('Deploying Training Function ')
for x in variables_categoricas:
df[x] = df[x].astype(str)
for x in variables_binarias:
df[x] = df[x].fillna(0)
df[x] = df[x].astype(np.int8)
for x in variables_numericas:
df[x] = df[x].apply(lambda c: getCleanPrice(c))
df[x] = df[x].astype(float)
full_pipeline.transform(df)
tree_predictions = tree_reg.predict(listing_prepared)
tree_mse = mean_squared_error(listing_label, tree_predictions)
tree_rmse = np.sqrt(tree_mse)
tree_mae = mean_absolute_error(listing_label, tree_predictions)
print("DecisionTreeRegressor:\nRMSE = ${:.4f}\nMAE = ${:.4f}".format(
tree_rmse, tree_mae))
print("-" * 100)
forest_predictions = forest_reg.predict(listing_prepared)
forest_mse = mean_squared_error(listing_label, forest_predictions)
forest_rmse = np.sqrt(forest_mse)
forest_mae = mean_absolute_error(listing_label, forest_predictions)
print("RandomForestRegressor:\nRMSE = ${:.4f}\nMAE = ${:.4f}".format(
forest_rmse, forest_mae))
return [price for price in forest_predictions]
cliente = Client("http://localhost:9004")
cliente.deploy("hostPricePredictor",
hostPricePredictor,
"Calcula el precio del hosting segun los parametros recibidos",
override=True)