modified: experiments.py

	modified:   models/cmhe/__init__.py
	modified:   models/cmhe/cmhe_torch.py
	modified:   models/dcm/__init__.py
	modified:   phenotyping.py
	modified:   reporting.py
	new file:   ../examples/.ipynb_checkpoints/Demo of CMHE on Synthetic Data-checkpoint.ipynb
	modified:   ../examples/Demo of CMHE on Synthetic Data.ipynb
	new file:   ../examples/Phenotyping Censored Time-to-Events.ipynb
	new file:   ../examples/Survival Regression with Auton-Survival.ipynb
	new file:   ../examples/__pycache__/cmhe_demo_utils.cpython-38.pyc
	new file:   ../examples/__pycache__/estimators_demo_utils.cpython-38.pyc
	modified:   ../examples/cmhe_demo_utils.py
	new file:   ../examples/scratch.ipynb

.pylintrc

@@ -23,7 +23,7 @@ persistent=no
 load-plugins=
 
 # Use multiple processes to speed up Pylint.
-jobs=4
+jobs=16
 
 # Allow loading of arbitrary C extensions. Extensions are imported into the
 # active Python interpreter and may run arbitrary code.

.vscode/settings.json

@@ -0,0 +1,4 @@
+{
+	"python.linting.pylintEnabled": true,
+	"python.linting.enabled": true
+}

auton_survival/experiments.py

@@ -63,7 +63,8 @@ def __init__(self, model, cv_folds=5, random_seed=0, hyperparam_grid={}):
 
   def fit(self, features, outcomes, ret_trained_model=True):
 
-    r"""Fits the Survival Regression Model to the data in a Cross Validation fashion.
+    r"""Fits the Survival Regression Model to the data in a Cross
+    Validation fashion.
 
     Parameters
     -----------
@@ -76,8 +77,8 @@ def fit(self, features, outcomes, ret_trained_model=True):
         a column named 'event' that contains the censoring status.
         \( \delta_i = 1 \) if the event is observed.
     ret_trained_model : bool
-        If True, the trained model is returned. If False, the fit function returns
-        self.
+        If True, the trained model is returned. If False, the fit function
+        returns self.
 
     Returns
     -----------

auton_survival/models/cmhe/__init__.py

@@ -116,12 +116,17 @@ class DeepCoxMixturesHeterogenousEffects:
 
   """
 
-  def __init__(self, k, g, layers=None):
+  def __init__(self, k, g, layers=None, gamma=100,
+               smoothing_factor=1e-4,
+               random_seed=0):
 
-    self.layers = layers
-    self.fitted = False
     self.k = k
     self.g = g
+    self.layers = layers
+    self.fitted = False
+    self.gamma = gamma
+    self.smoothing_factor = smoothing_factor
+    self.random_seed = random_seed
 
   def __call__(self):
     if self.fitted:
@@ -176,8 +181,14 @@ def _preprocess_training_data(self, x, t, e, a, vsize, val_data,
 
   def _gen_torch_model(self, inputdim, optimizer):
     """Helper function to return a torch model."""
+
+    np.random.seed(self.random_seed)
+    torch.manual_seed(self.random_seed)
+
     return DeepCMHETorch(self.k, self.g, inputdim,
                          layers=self.layers,
+                         gamma=self.gamma,
+                         smoothing_factor=self.smoothing_factor,
                          optimizer=optimizer)
 
   def fit(self, x, t, e, a, vsize=0.15, val_data=None,
@@ -235,7 +246,9 @@ def fit(self, x, t, e, a, vsize=0.15, val_data=None,
                           lr=learning_rate,
                           bs=batch_size,
                           patience=patience,
-                          return_losses=True)
+                          return_losses=True,
+                          use_posteriors=True,
+                          random_seed=self.random_seed)
 
     self.torch_model = (model[0].eval(), model[1])
     self.fitted = True

auton_survival/models/cmhe/cmhe_torch.py

@@ -43,7 +43,8 @@ def _init_dcmhe_layers(self, lastdim):
     self.phi_gate = torch.nn.Linear(lastdim, self.g, bias=False)
     self.omega = torch.nn.Parameter(torch.rand(self.g)-0.5)
 
-  def __init__(self, k, g, inputdim, layers=None, optimizer='Adam'):
+  def __init__(self, k, g, inputdim, layers=None, gamma=100,
+               smoothing_factor=1e-4, optimizer='Adam'):
 
     super(DeepCMHETorch, self).__init__()
 
@@ -56,6 +57,9 @@ def __init__(self, k, g, inputdim, layers=None, optimizer='Adam'):
     self.k = k # Base Physiology groups
     self.g = g # Treatment Effect groups
 
+    self.gamma = gamma
+    self.smoothing_factor = smoothing_factor
+
     if len(layers) == 0: lastdim = inputdim
     else: lastdim = layers[-1]
 
@@ -69,7 +73,9 @@ def forward(self, x, a):
     x = self.embedding(x)
     a = 2*(a-0.5)
 
-    log_hrs = torch.clamp(self.expert(x), min=-100, max=100)
+    log_hrs = torch.clamp(self.expert(x),
+                          min=-self.gamma,
+                          max=self.gamma)
 
     logp_z_gate = torch.nn.LogSoftmax(dim=1)(self.z_gate(x)) #
     logp_phi_gate = torch.nn.LogSoftmax(dim=1)(self.phi_gate(x))
@@ -87,15 +93,3 @@ def forward(self, x, a):
         logp_joint_hrs[:, i, j] = log_hrs[:, i] + (j!=2)*a*self.omega[j]
 
     return logp_jointlatent_gate, logp_joint_hrs
-
-# class DeepCoxMixtureHETorch(CoxMixtureHETorch):
-
-#   def __init__(self, k, g, inputdim, hidden):
-
-#     super(DeepCoxMixtureHETorch, self).__init__(k, g, inputdim, hidden)
-
-#     # Get rich feature representations of the covariates
-#     self.embedding = torch.nn.Sequential(torch.nn.Linear(inputdim, hidden),
-#                                          torch.nn.Tanh(),
-#                                          torch.nn.Linear(hidden, hidden),
-#                                          torch.nn.Tanh())

auton_survival/models/dcm/__init__.py

@@ -79,6 +79,12 @@ class DeepCoxMixtures:
   layers: list
       A list of integers consisting of the number of neurons in each
       hidden layer.
+<<<<<<< Updated upstream
+=======
+  random_seed: int
+      Controls the reproducibility of called functions.
+ 
+>>>>>>> Stashed changes
   Example
   -------
   >>> from auton_survival.models.dcm import DeepCoxMixtures

auton_survival/phenotyping.py

@@ -320,8 +320,9 @@ def fit(self, features):
 
     """
 
-    if self.dim_red_method is not None: 
-      print("Fitting the following Dimensionality Reduction Model:\n", self.dim_red_model)
+    if self.dim_red_method is not None:
+      print("Fitting the following Dimensionality Reduction Model:\n", 
+            self.dim_red_model)
       self.dim_red_model = self.dim_red_model.fit(features)
       features = self.dim_red_model.transform(features)
 
@@ -457,8 +458,9 @@ def fit(self, features, outcomes, interventions, horizon):
     self.cf_model = cf_model.fit(features, outcomes, interventions) 
 
     times = np.unique(outcomes.times.values)
-    cf_predictions = self.cf_model.predict_counterfactual_survival(features, interventions, times)
-
+    cf_predictions = self.cf_model.predict_counterfactual_survival(features,
+                                                                   times)
+
     ite_estimates = cf_predictions[1] - cf_predictions[0]
 
     if self.phenotyping_method == 'rsf':

auton_survival/reporting.py

@@ -3,9 +3,7 @@
 
 from lifelines import KaplanMeierFitter, NelsonAalenFitter
 
-from collections import Counter
-
-from lifelines import KaplanMeierFitter, CoxPHFitter
+from lifelines import KaplanMeierFitter
 from lifelines.plotting import add_at_risk_counts
 
 
@@ -75,10 +73,10 @@ def plot_nelsonaalen(outcomes, groups=None, **kwargs):
   for group in sorted(set(groups)):
     if pd.isna(group): continue
 
-    print("Group:", group)
+    print('Group:', group)
 
-    NelsonAalenFitter().fit(outcomes[groups==group]['time'], 
-                            outcomes[groups==group]['event']).plot(label=group, 
+    NelsonAalenFitter().fit(outcomes[groups==group]['time'],
+                            outcomes[groups==group]['event']).plot(label=group,
                                                                    **kwargs)