def _ada_boost_regression_train(table,
feature_cols,
label_col,
max_depth=3,
n_estimators=50,
learning_rate=1.0,
loss='linear',
random_state=None):
feature_names, x_train = check_col_type(table, feature_cols)
y_train = table[label_col]
base_estimator = DecisionTreeRegressor(max_depth=max_depth)
regressor = AdaBoostRegressor(base_estimator, n_estimators, learning_rate,
loss, random_state)
regressor.fit(x_train, y_train)
params = {
'feature_cols': feature_cols,
'label_col': label_col,
'feature_importance': regressor.feature_importances_,
'n_estimators': n_estimators,
'learning_rate': learning_rate,
'loss': loss,
'random_state': random_state
}
model = _model_dict('ada_boost_regression_model')
get_param = regressor.get_params()
model['parameters'] = get_param
model['regressor'] = regressor
model['params'] = params
fig_feature_importance = _plot_feature_importance(feature_names, regressor)
params = dict2MD(get_param)
rb = BrtcReprBuilder()
rb.addMD(
strip_margin("""
| ## AdaBoost Regression Train Result
|
| ### Feature Importance
| {fig_feature_importance}
|
| ### Parameters
| {list_parameters}
|
""".format(fig_feature_importance=fig_feature_importance,
list_parameters=params)))
model['_repr_brtc_'] = rb.get()
feature_importance = regressor.feature_importances_
feature_importance_table = pd.DataFrame(
[[feature_names[i], feature_importance[i]]
for i in range(len(feature_names))],
columns=['feature_name', 'importance'])
model['feature_importance_table'] = feature_importance_table
return {'model': model}
def test_parameters(self):
""" Testing parameters of Model class. """
#1.)
#create instance of PLS model using Model class & creating instance
# using SKlearn libary, comparing if the parameters of both instances are equal
pls_parameters = {"n_components": 20, "scale": False, "max_iter": 200}
model = Model(algorithm="PlsRegression", parameters=pls_parameters)
pls_model = PLSRegression(n_components=20, scale="svd", max_iter=200)
for k, v in model.model.get_params().items():
self.assertIn(k, list(pls_model.get_params()))
#2.)
rf_parameters = {"n_estimators": 200, "max_depth": 50,"min_samples_split": 10}
model = Model(algorithm="RandomForest", parameters=rf_parameters)
rf_model = RandomForestRegressor(n_estimators=200, max_depth=50, min_samples_split=10)
for k, v in model.model.get_params().items():
self.assertIn(k, list(rf_model.get_params()))
#3.)
knn_parameters = {"n_neighbors": 10, "weights": "distance", "algorithm": "ball_tree"}
model = Model(algorithm="KNN", parameters=knn_parameters)
knn_model = KNeighborsRegressor(n_neighbors=10, weights='distance', algorithm="kd_tree")
for k, v in model.model.get_params().items():
self.assertIn(k, list(knn_model.get_params()))
#4.)
svr_parameters = {"kernel": "poly", "degree": 5, "coef0": 1}
model = Model(algorithm="SVR",parameters=svr_parameters)
svr_model = SVR(kernel='poly', degree=5, coef0=1)
for k, v in model.model.get_params().items():
self.assertIn(k, list(svr_model.get_params()))
#5.)
ada_parameters = {"n_estimators": 150, "learning_rate": 1.2, "loss": "square"}
model = Model(algorithm="AdaBoost", parameters=ada_parameters)
ada_model = AdaBoostRegressor(n_estimators=150, learning_rate=1.2, loss="square")
for k, v in model.model.get_params().items():
self.assertIn(k, list(ada_model.get_params()))
#6.)
bagging_parameters = {"n_estimators": 50, "max_samples": 1.5, "max_features": 2}
model = Model(algorithm="Bagging", parameters=bagging_parameters)
bagging_model = BaggingRegressor(n_estimators=50, max_samples=1.5, max_features="square")
for k, v in model.model.get_params().items():
self.assertIn(k, list(bagging_model.get_params()))
#7.)
lasso_parameters = {"alpha": 1.5, "max_iter": 500, "tol": 0.004}
model = Model(algorithm="lasso", parameters=lasso_parameters)
lasso_model = Lasso(alpha=1.5, max_iter=500, tol=0.004)
for k, v in model.model.get_params().items():
self.assertIn(k, list(lasso_model.get_params()))
class _AdaBoostRegressorImpl:
def __init__(
self,
base_estimator=None,
*,
n_estimators=50,
learning_rate=1.0,
loss="linear",
random_state=None,
):
if base_estimator is None:
estimator_impl = None
else:
estimator_impl = _FitSpecProxy(base_estimator)
self._hyperparams = {
"base_estimator": estimator_impl,
"n_estimators": n_estimators,
"learning_rate": learning_rate,
"loss": loss,
"random_state": random_state,
}
self._wrapped_model = SKLModel(**self._hyperparams)
self._hyperparams["base_estimator"] = base_estimator
def get_params(self, deep=True):
out = self._wrapped_model.get_params(deep=deep)
# we want to return the lale operator, not the underlying impl
out["base_estimator"] = self._hyperparams["base_estimator"]
return out
def fit(self, X, y=None):
if isinstance(X, pd.DataFrame):
feature_transformer = FunctionTransformer(
func=lambda X_prime: pd.DataFrame(X_prime, columns=X.columns),
inverse_func=None,
check_inverse=False,
)
self._hyperparams["base_estimator"] = _FitSpecProxy(
feature_transformer >> self._hyperparams["base_estimator"])
self._wrapped_model = SKLModel(**self._hyperparams)
if y is not None:
self._wrapped_model.fit(X, y)
else:
self._wrapped_model.fit(X)
return self
def predict(self, X):
return self._wrapped_model.predict(X)
def score(self, X, y, sample_weight=None):
return self._wrapped_model.score(X, y, sample_weight)
class _AdaBoostRegressorImpl:
def __init__(
self,
base_estimator=None,
n_estimators=50,
learning_rate=1.0,
loss="linear",
random_state=None,
):
estimator_impl = base_estimator
if isinstance(estimator_impl, lale.operators.Operator):
if isinstance(estimator_impl, lale.operators.IndividualOp):
estimator_impl = estimator_impl._impl_instance()
wrapped_model = getattr(estimator_impl, "_wrapped_model", None)
if wrapped_model is not None:
estimator_impl = wrapped_model
else:
raise ValueError(
"If base_estimator is a Lale operator, it needs to be an individual operator. "
)
self._hyperparams = {
"base_estimator": estimator_impl,
"n_estimators": n_estimators,
"learning_rate": learning_rate,
"loss": loss,
"random_state": random_state,
}
self._wrapped_model = SKLModel(**self._hyperparams)
self._hyperparams["base_estimator"] = base_estimator
def get_params(self, deep=True):
out = self._wrapped_model.get_params(deep=deep)
# we want to return the lale operator, not the underlying impl
out["base_estimator"] = self._hyperparams["base_estimator"]
return out
def fit(self, X, y=None):
if y is not None:
self._wrapped_model.fit(X, y)
else:
self._wrapped_model.fit(X)
return self
def predict(self, X):
return self._wrapped_model.predict(X)
def score(self, X, y, sample_weight=None):
return self._wrapped_model.score(X, y, sample_weight)
def ada_boost(df, significant_cols, target, cat_cols, num_cols):
ss = StandardScaler()
ohe = OneHotEncoder(drop='first', sparse=False)
X = df[significant_cols]
y = df[target]
base = DecisionTreeRegressor(max_depth=3, random_state=0)
estimator = AdaBoostRegressor(base_estimator=base, random_state=0)
params = {
'n_estimators': np.arange(5, int(X.shape[0] * 0.1)),
'learning_rate': np.arange(0.1, 1.1, 0.1),
'loss': ['linear', 'square', 'exponential'],
}
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.1,
random_state=0)
X_train_cat = ohe.fit_transform(X_train[cat_cols])
X_train_num = ss.fit_transform(X_train[num_cols])
X_test_cat = ohe.transform(X_test[cat_cols])
X_test_num = ss.transform(X_test[num_cols])
train_data = np.c_[X_train_cat, X_train_num]
test_data = np.c_[X_test_cat, X_test_num]
gs = GridSearchCV(estimator, params, scoring='r2', cv=3)
gs.fit(train_data, y_train)
estimator = gs.best_estimator_
r2_cv_scores = cross_val_score(estimator,
train_data,
y_train,
scoring='r2',
cv=3,
n_jobs=-1)
rmse_cv_scores = cross_val_score(estimator,
train_data,
y_train,
scoring='neg_root_mean_squared_error',
cv=3,
n_jobs=-1)
params = estimator.get_params()
r2 = np.mean(r2_cv_scores)
rmse = np.abs(np.mean(rmse_cv_scores))
r2_variance = np.var(r2_cv_scores, ddof=1)
rmse_variance = np.abs(np.var(rmse_cv_scores, ddof=1))
estimator.fit(train_data, y_train)
y_pred = estimator.predict(test_data)
r2_validation = r2_score(y_test, y_pred)
rmse_validation = np.sqrt(mean_squared_error(y_test, y_pred))
return r2, rmse, r2_variance, rmse_variance, r2_validation, rmse_validation, params
def ada(X, Y, kfold=3, feature_set=None):
arr = index_splitter(N=len(X), fold=kfold)
ps = PredefinedSplit(arr)
for train, test in ps.split():
train_index = train
test_index = test
train_X, train_y = X.values[train_index, :], Y.values[train_index]
test_X, test_y = X.values[test_index, :], Y.values[test_index]
arr = index_splitter(N=len(train_X), fold=kfold)
ps2 = PredefinedSplit(arr)
learning_rate = [x for x in np.linspace(0.1, 1, num=10)]
n_estimators = [int(x) for x in np.linspace(start=20, stop=1000, num=100)]
loss = ['square']
random_grid = {
'n_estimators': n_estimators,
'learning_rate': learning_rate,
'loss': loss
}
# Use the random grid to search for best hyperparameters
# First create the base model to tune
ada = AdaBoostRegressor(random_state=42, loss='square')
# Look at parameters used by our current forest
print('Parameters for baseline:\n')
pprint(ada.get_params())
# Random search of parameters, using 3 fold cross validation,
# search across 100 different combinations, and use all available cores
ada_random = RandomizedSearchCV(estimator=ada,
n_iter=200,
param_distributions=random_grid,
scoring='neg_mean_squared_error',
cv=ps2.split(),
verbose=2,
random_state=42,
n_jobs=-1)
# Fit the random search model
ada_random.fit(train_X, train_y)
pprint(ada_random.best_params_)
cv_result_rd = ada_random.cv_results_
BestPara_random = ada_random.best_params_
## Grid search of parameters, using 3 fold cross validation based on Random search
lr = [BestPara_random['learning_rate']]
#n_estimators = [BestPara_random["n_estimators"]]
n_estimators = [
int(x) for x in range(BestPara_random["n_estimators"] -
10, BestPara_random["n_estimators"] + 10, 20)
]
n_estimators = [item for item in n_estimators if item > 0]
grid_grid = {
'n_estimators': n_estimators,
'learning_rate': lr,
'loss': loss
}
ada_grid = GridSearchCV(estimator=ada,
param_grid=grid_grid,
scoring='neg_mean_squared_error',
cv=ps2.split(),
verbose=2,
n_jobs=-1)
# Fit the grid search model
ada_grid.fit(train_X, train_y)
BestPara_grid = ada_grid.best_params_
pprint(ada_grid.best_params_)
cv_results_grid = ada_grid.cv_results_
# Fit the base line search model
ada.fit(train_X, train_y)
#prediction
predict_y = ada_random.predict(test_X)
predict_y_grid = ada_grid.predict(test_X)
predict_y_base = ada.predict(test_X)
# Performance metrics
def RMLSE(predict_y_grid, predict_y, predict_y_base, test_y):
errors_Grid_CV = np.sqrt(mean_squared_log_error(
predict_y_grid, test_y))
errors_Random_CV = np.sqrt(mean_squared_log_error(predict_y, test_y))
errors_baseline = np.sqrt(
mean_squared_log_error(predict_y_base, test_y))
return errors_Grid_CV, errors_Random_CV, errors_baseline
errors_Grid_CV = (mean_squared_error(predict_y_grid,
test_y)) #,squared = False))
errors_Random_CV = (mean_squared_error(predict_y,
test_y)) #,squared = False))
errors_baseline = (mean_squared_error(predict_y_base,
test_y)) #,squared = False))
x_axis = range(3)
results = [errors_Grid_CV, errors_Random_CV, errors_baseline]
print('Adaboot Results:', results)
if True:
fig = plt.figure(figsize=(15, 8))
x_axis = range(3)
plt.bar(x_axis, results)
plt.xticks(x_axis, ('GridSearchCV', 'RandomizedSearchCV', 'Baseline'))
#plt.show()
plt.savefig('ada_compare_error.png')
#feature importance
num_feature = len(ada_grid.best_estimator_.feature_importances_)
plt.figure(figsize=(24, 6))
plt.bar(range(0, num_feature * 4, 4),
ada_grid.best_estimator_.feature_importances_)
label_name = X.keys()
plt.xticks(range(0, num_feature * 4, 4), label_name)
plt.title("Feature Importances" + ",kfold=" + str(kfold))
#plt.show()
plt.savefig('ada_feature_importance.png')
fig = plt.figure(figsize=(20, 8))
ax = fig.gca()
x_label = range(0, len(predict_y_grid))
plt.title("kfold=" + str(kfold))
ax.plot(x_label, predict_y_grid, 'r--', label="predict")
ax.plot(x_label, test_y, label="ground_truth")
ax.set_ylim(0, 200)
ax.legend()
#plt.show()
plt.savefig('ada_prediction.png')
#return a dictionary for all results
return ada_grid.predict, ada_grid.best_estimator_
y = yacht["resid_resist"]
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=123)
# MinMaxScaler da mejores resultados que StanderScaler
scaler = MinMaxScaler()
train_scaled = scaler.fit_transform(X_train)
test_scaled = scaler.transform(X_test)
model = AdaBoostRegressor(base_estimator=RandomForestRegressor())
model.fit(train_scaled, y_train)
print("Accuracy on train data: ", round(model.score(train_scaled, y_train)*100, 2), "%")
print("Accuracy on test data: ", round(model.score(test_scaled, y_test)*100, 2), "%")
print("Parameters: ", model.get_params())
print("MAE: ", mean_absolute_error(y_test, model.predict(test_scaled)))
# TODO: Se puede mejorar el Grid
gridParams = {
"n_estimators": [200], 'base_estimator__n_estimators': np.arange(1, 20)}
grid = GridSearchCV(model, gridParams,
verbose=1,
cv=5)
grid.fit(train_scaled, y_train)
print("Best params:", grid.best_params_)
print("Best score:", grid.best_score_)
params = grid.best_params_
y = np.sin(X1).ravel() + np.sin(6 * X1).ravel() + rng.normal(
0, 0.1, X1.shape[0])
print(X.shape, y.shape)
# Fit regression model
regr_1 = DecisionTreeRegressor(max_depth=4)
regr_2 = AdaBoostRegressor(DecisionTreeRegressor(max_depth=4),
n_estimators=300,
random_state=rng)
regr_1.fit(X, y)
regr_2.fit(X, y)
# Predict
y_1 = regr_1.predict(X)
y_2 = regr_2.predict(X)
print(regr_2.get_params())
# Plot the results
# plt.figure()
# plt.scatter(X1, y, c="k", label="training samples")
# plt.plot(X1, y_1, c="g", label="n_estimators=1", linewidth=2)
# plt.plot(X1, y_2, c="r", label="n_estimators=300", linewidth=2)
# plt.xlabel("data")
# plt.ylabel("target")
# plt.title("Boosted Decision Tree Regression")
# plt.legend()
# plt.show()