Commit ee24ea41 authored by Bharath Ramsundar's avatar Bharath Ramsundar
Browse files

Cleanup and bug fixes

parent f5c9223f

deepchem/models/multitask.py

+1 −1
Original line number Diff line number Diff line
@@ -65,7 +65,7 @@ class SingletaskToMultitask(Model): 
    n_samples = X.shape[0]

    y_pred = np.zeros((n_samples, n_tasks))

    for ind, task in enumerate(self.tasks):

      task_type = task_types[task]

      task_type = self.task_types[task]

      if task_type == "classification":

        y_pred[:, ind] = self.models[task].predict_on_batch(X)[:, 0]

      elif task_type == "regression":

deepchem/models/sklearn_models/__init__.py

+0 −5
Original line number Diff line number Diff line
@@ -43,11 +43,6 @@ class SklearnModel(Model): 
    """

    X, y, w, _ = dataset.to_numpy()

    y, w = np.squeeze(y), np.squeeze(w)

    ######## DEBUG

    print("SklearnModel.fit()")

    print("X.shape, y.shape, w.shape")

    print(X.shape, y.shape, w.shape)

    ######## DEBUG

    self.raw_model.fit(X, y, w)

    y_pred_raw = self.raw_model.predict(X)

deepchem/models/tensorflow_models/__init__.py

+0 −6
Original line number Diff line number Diff line
@@ -139,12 +139,6 @@ class TensorflowGraph(object): 
      gradient_costs = []  # costs used for gradient calculation



      with self._shared_name_scope('costs'):

        ######## DEBUG

        print("self.num_tasks")

        print(self.num_tasks)

        print("len(self.output)")

        print(len(self.output))

        ######## DEBUG

        for task in xrange(self.num_tasks):

          task_str = str(task).zfill(len(str(self.num_tasks)))

          with self._shared_name_scope('cost_{}'.format(task_str)):

deepchem/models/tensorflow_models/model_ops.py

+0 −7
Original line number Diff line number Diff line
@@ -300,13 +300,6 @@ def MultitaskLogits(features, num_tasks, num_classes=2, weight_init=None, 
        logits.append(

            Logits(features, num_classes, weight_init=weight_init,

                   bias_init=bias_init, dropout=dropout))

  ###### DEBUG

  print("MultitaskLogits")

  print("num_tasks")

  print(num_tasks)

  print("len(logits)")

  print(len(logits))

  ###### DEBUG

  return logits

deepchem/models/tests/test_overfit.py

+0 −60
Original line number Diff line number Diff line
@@ -143,9 +143,6 @@ class TestOverfitAPI(TestAPI): 
    model.fit(dataset)

    model.save()



    y_pred_model = model.predict(dataset, transformers=[])

    y_pred_proba_model = model.predict_proba(dataset, transformers=[])



    # Eval model on train

    transformers = []

    evaluator = Evaluator(model, dataset, transformers, verbosity=verbosity)
@@ -195,8 +192,6 @@ class TestOverfitAPI(TestAPI): 
    model.fit(dataset)

    model.save()



    y_pred_model = model.predict(dataset, transformers=[])



    # Eval model on train

    transformers = []

    evaluator = Evaluator(model, dataset, transformers, verbosity=verbosity)
@@ -248,8 +243,6 @@ class TestOverfitAPI(TestAPI): 
    model.fit(dataset)

    model.save()



    y_pred_model = model.predict(dataset, transformers=[])



    # Eval model on train

    transformers = []

    evaluator = Evaluator(model, dataset, transformers, verbosity=verbosity)
@@ -299,9 +292,6 @@ class TestOverfitAPI(TestAPI): 
    model.fit(dataset)

    model.save()



    y_pred_model = model.predict(dataset, transformers=[])

    y_pred_proba_model = model.predict_proba(dataset, transformers=[])



    # Eval model on train

    transformers = []

    evaluator = Evaluator(model, dataset, transformers, verbosity=verbosity)
@@ -352,9 +342,6 @@ class TestOverfitAPI(TestAPI): 
    model.fit(dataset)

    model.save()



    y_pred_model = model.predict(dataset, transformers=[])

    y_pred_proba_model = model.predict_proba(dataset, transformers=[])



    # Eval model on train

    transformers = []

    evaluator = Evaluator(model, dataset, transformers, verbosity=verbosity)
@@ -409,9 +396,6 @@ class TestOverfitAPI(TestAPI): 
    model.fit(dataset)

    model.save()



    y_pred_model = model.predict(dataset, transformers=[])

    y_pred_proba_model = model.predict_proba(dataset, transformers=[])



    # Eval model on train

    transformers = []

    evaluator = Evaluator(model, dataset, transformers, verbosity=verbosity)
@@ -467,9 +451,6 @@ class TestOverfitAPI(TestAPI): 
    model.fit(dataset)

    model.save()



    y_pred_model = model.predict(dataset, transformers=[])

    y_pred_proba_model = model.predict_proba(dataset, transformers=[])



    # Eval model on train

    transformers = []

    evaluator = Evaluator(model, dataset, transformers, verbosity=verbosity)
@@ -487,25 +468,13 @@ class TestOverfitAPI(TestAPI): 
    

    tasks = ["task0"]

    task_types = {task: "classification" for task in tasks}

    #n_samples = 250

    #n_samples = 500

    #n_samples = 1000

    #n_samples = 2000

    #n_samples = 5000

    n_samples = 5120

    #n_features = 3

    n_features = 6

    n_tasks = len(tasks)

    n_classes = 2

    

    # Generate dummy dataset

    np.random.seed(123)

    #p = .002

    #p = .2

    #p = .1

    #p = .05

    #p = .01

    #p = .005

    p = .002

    ids = np.arange(n_samples)

    X = np.random.rand(n_samples, n_features)
@@ -513,14 +482,6 @@ class TestOverfitAPI(TestAPI): 
    w = np.ones((n_samples, n_tasks))

    print("np.count_nonzero(y)")

    print(np.count_nonzero(y))

    #w = np.random.binomial(1, p, size=(n_samples, n_tasks))

    #print("np.amin(w), np.amax(w)")

    #print(np.amin(w), np.amax(w))



    #print("y_nonzero.shape")

    #print(y_nonzero.shape)

    #print("np.count_nonzero(y_nonzero)")

    #print(np.count_nonzero(y_nonzero))

    ##### DEBUG

    y_flat, w_flat = np.squeeze(y), np.squeeze(w)

    y_nonzero = y_flat[w_flat != 0]
@@ -541,17 +502,7 @@ class TestOverfitAPI(TestAPI): 
      "dropouts": [.0],

      "learning_rate": 0.003,

      "momentum": .9,

      #"batch_size": n_samples,

      #"batch_size": n_samples/2,

      #"batch_size": n_samples/4,

      #"batch_size": n_samples/8,

      #"batch_size": n_samples/16,

      #"batch_size": n_samples/32,

      #"batch_size": n_samples/64,

      "batch_size": 75,

      # TODO(rbharath): Is there a bug in the padding code? Why does it fail to

      # learn for non-multiples?

      #"batch_size": 600,

      "num_classification_tasks": 1,

      "num_classes": n_classes,

      "num_features": n_features,
@@ -562,8 +513,6 @@ class TestOverfitAPI(TestAPI): 
      "optimizer": "adam",

      "data_shape": dataset.get_data_shape()

    }

    print("model_params['batch_size']")

    print(model_params['batch_size'])



    verbosity = "high"

    classification_metric = Metric(metrics.roc_auc_score, verbosity=verbosity)
@@ -576,9 +525,6 @@ class TestOverfitAPI(TestAPI): 
    model.fit(dataset)

    model.save()



    y_pred_model = model.predict(dataset, transformers=[])

    y_pred_proba_model = model.predict_proba(dataset, transformers=[])



    # Eval model on train

    transformers = []

    evaluator = Evaluator(model, dataset, transformers, verbosity=verbosity)
@@ -671,9 +617,6 @@ class TestOverfitAPI(TestAPI): 
    model.fit(dataset)

    model.save()



    y_pred_model = model.predict(dataset, transformers=[])

    y_pred_proba_model = model.predict_proba(dataset, transformers=[])



    # Eval model on train

    transformers = []

    evaluator = Evaluator(model, dataset, transformers, verbosity=verbosity)
@@ -728,9 +671,6 @@ class TestOverfitAPI(TestAPI): 
    model.fit(dataset)

    model.save()



    y_pred_model = model.predict(dataset, transformers=[])

    y_pred_proba_model = model.predict_proba(dataset, transformers=[])



    # Eval model on train

    transformers = []

    evaluator = Evaluator(model, dataset, transformers, verbosity=verbosity)

CRA Git | Maintained and supported by SUSTech CRA and CCSE