Skip to content

Commit

Permalink
fix tests (#284)
Browse files Browse the repository at this point in the history
  • Loading branch information
@pplonski
pplonski committed Jan 10, 2021
1 parent 3df2cc1 commit da293b0
Show file tree
Hide file tree
Showing 2 changed files with 44 additions and 20 deletions.
61 changes: 41 additions & 20 deletions supervised/tuner/mljar_tuner.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -138,12 +138,14 @@ def generate_params(
elif step == "loo_encoding":
return self.get_loo_categorical_strategy(models, total_time_limit)
elif step == "golden_features":
return self.get_golden_features_params(models, results_path, total_time_limit)
return self.get_golden_features_params(
models, results_path, total_time_limit
)
elif step == "insert_random_feature":
return self.get_params_to_insert_random_feature(models, total_time_limit)
elif step == "features_selection":
return self.get_features_selection_params(
self.filter_random_feature_model(models), results_path
self.filter_random_feature_model(models), results_path, total_time_limit
)
elif "hill_climbing" in step:
return self.get_hill_climbing_params(
Expand Down Expand Up @@ -484,7 +486,9 @@ def get_loo_categorical_strategy(self, current_models, total_time_limit):

def get_categorical_strategy(self, current_models, strategy, total_time_limit):

df_models, algorithms = self.df_models_algorithms(current_models, time_limit=0.03*total_time_limit)
df_models, algorithms = self.df_models_algorithms(
current_models, time_limit=0.03 * total_time_limit
)
generated_params = []
for m_type in algorithms:
# try to add categorical strategy only for below algorithms
Expand Down Expand Up @@ -548,9 +552,15 @@ def df_models_algorithms(self, current_models, time_limit=None):
model_types = [m.get_type() for m in current_models]
names = [m.get_name() for m in current_models]
train_times = [m.get_train_time() for m in current_models]

df_models = pd.DataFrame(
{"model": current_models, "score": scores, "model_type": model_types, "name": names, "train_time": train_times}
{
"model": current_models,
"score": scores,
"model_type": model_types,
"name": names,
"train_time": train_times,
}
)
if time_limit is not None:
df_models = df_models[df_models.train_time < time_limit]
Expand All @@ -560,15 +570,18 @@ def df_models_algorithms(self, current_models, time_limit=None):
model_types = list(df_models.model_type)
u, idx = np.unique(model_types, return_index=True)
algorithms = u[np.argsort(idx)]

#print(df_models)
#print(algorithms)
return df_models, algorithms

# print(df_models)
# print(algorithms)
return df_models, algorithms

def get_golden_features_params(self, current_models, results_path, total_time_limit):
def get_golden_features_params(
self, current_models, results_path, total_time_limit
):

df_models, algorithms = self.df_models_algorithms(current_models, time_limit=0.03*total_time_limit)
df_models, algorithms = self.df_models_algorithms(
current_models, time_limit=0.03 * total_time_limit
)

generated_params = []
for m_type in algorithms:
Expand All @@ -579,7 +592,7 @@ def get_golden_features_params(self, current_models, results_path, total_time_li

for i in range(min(1, len(models))):
m = models.iloc[i]

params = copy.deepcopy(m.params)
params["preprocessing"]["golden_features"] = {
"results_path": results_path,
Expand All @@ -598,9 +611,11 @@ def get_golden_features_params(self, current_models, results_path, total_time_li
generated_params += [params]
return generated_params

def time_features_selection(self, current_models):

df_models, algorithms = self.df_models_algorithms(current_models)
def time_features_selection(self, current_models, total_time_limit):

df_models, algorithms = self.df_models_algorithms(
current_models, time_limit=0.05 * total_time_limit
)

time_needed = 0
for m_type in algorithms:
Expand Down Expand Up @@ -631,7 +646,7 @@ def time_features_selection(self, current_models):

def get_params_to_insert_random_feature(self, current_models, total_time_limit):

time_needed = self.time_features_selection(current_models)
time_needed = self.time_features_selection(current_models, total_time_limit)

if time_needed > 0.1 * total_time_limit:
print("Not enough time to perform features selection. Skip")
Expand All @@ -640,7 +655,9 @@ def get_params_to_insert_random_feature(self, current_models, total_time_limit):
)
return None

df_models, algorithms = self.df_models_algorithms(current_models)
df_models, algorithms = self.df_models_algorithms(
current_models, time_limit=0.05 * total_time_limit
)

m = df_models.iloc[0]["model"]

Expand All @@ -660,7 +677,9 @@ def get_params_to_insert_random_feature(self, current_models, total_time_limit):
return [params]
return None

def get_features_selection_params(self, current_models, results_path):
def get_features_selection_params(
self, current_models, results_path, total_time_limit
):

fname = os.path.join(results_path, "drop_features.json")
if not os.path.exists(fname):
Expand All @@ -674,8 +693,10 @@ def get_features_selection_params(self, current_models, results_path):
# skip this step
if len(drop_features) <= 1:
return None

df_models, algorithms = self.df_models_algorithms(current_models)

df_models, algorithms = self.df_models_algorithms(
current_models, time_limit=0.05 * total_time_limit
)

generated_params = []
for m_type in algorithms:
Expand Down
3 changes: 3 additions & 0 deletions tests/tests_tuner/test_hill_climbing.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -19,6 +19,9 @@ def get_type(self):
def get_final_loss(self):
return self.final_loss

def get_train_time(self):
return 0.1


class TunerHillClimbingTest(unittest.TestCase):
def test_hill_climbing(self):
Expand Down

0 comments on commit da293b0

Please sign in to comment.