Merge pull request #100 from antoinedemathelin/master

fix: Tensorflow, sklearn updates

Author: antoinedemathelin

.github/workflows/run-test.yml

@@ -18,6 +18,8 @@ jobs:
         exclude:
           - os: windows-latest
             python-version: 3.9
+          - os: ubuntu-latest
+            python-version: 3.6
     runs-on: ${{ matrix.os }}
     steps:
     - uses: actions/checkout@v2

adapt/base.py

@@ -14,6 +14,11 @@
 from sklearn.exceptions import NotFittedError
 from tensorflow.keras import Model
 from tensorflow.keras.wrappers.scikit_learn import KerasClassifier, KerasRegressor
+try:
+    from tensorflow.keras.optimizers.legacy import RMSprop
+except:
+    from tensorflow.keras.optimizers import RMSprop
+
 
 from adapt.utils import (check_estimator,
                          check_network,
@@ -1308,7 +1313,7 @@ def compile(self,
 
         if ((not "optimizer" in compile_params) or 
             (compile_params["optimizer"] is None)):
-            compile_params["optimizer"] = "rmsprop"
+            compile_params["optimizer"] = RMSprop()
         else:
             if optimizer is None:
                 if not isinstance(compile_params["optimizer"], str):

adapt/feature_based/_fa.py

@@ -92,9 +92,9 @@ class FA(BaseAdaptEstimator):
 
     Notes
     -----
-    FA can be used for multi-source DA by giving list of source data
-    for arguments Xs, ys of fit method : Xs = [Xs1, Xs2, ...],
-    ys = [ys1, ys2, ...]
+    FA can be used for multi-source DA by using the ``domains`` argument
+    in the ``fit`` or ``fit_transform`` method. An example is given
+    `[here] <https://github.com/adapt-python/adapt/issues/86>`
     """
     def __init__(self,
                  estimator=None,

adapt/parameter_based/_transfer_tree.py

@@ -456,7 +456,11 @@ def _force_coherence(self,rule,node=0,Translate=False,indexes_nodes=list(),drift
             if not coh:
                 if Translate :
                     if auto_drift:
-                        b_infs,b_sups = ut.bounds_rule(rule,self.estimator_.n_features_)
+                        try:
+                            n_feat = self.estimator_.n_features_
+                        except:
+                            n_feat = self.estimator_.n_features_in_
+                        b_infs,b_sups = ut.bounds_rule(rule, n_feat)
                         if non_coherent_sense == -1:
                             if b_sups[phi] == np.inf:
                                 self.updateSplit(node,phi,th+D_MARGIN)
@@ -817,8 +821,12 @@ def _strut(self,X_target_node,Y_target_node,node=0,no_prune_on_cl=False,cl_no_pr
         maj_class = np.argmax(self.estimator_.tree_.value[node, :].copy())
 
         if min_drift is None or max_drift is None:
-            min_drift = np.zeros(self.estimator_.n_features_)
-            max_drift = np.zeros(self.estimator_.n_features_)
+            try:
+                n_feat = self.estimator_.n_features_
+            except:
+                n_feat = self.estimator_.n_features_in_
+            min_drift = np.zeros(n_feat)
+            max_drift = np.zeros(n_feat)
 
         current_class_distribution = ut.compute_class_distribution(classes_, Y_target_node)
         is_reached = (Y_target_node.size > 0)

requirements.txt

@@ -1,5 +1,5 @@
 numpy
 scipy
-tensorflow >= 2.0
+tensorflow < 2.12
 scikit-learn
 cvxopt

tests/test_adda.py

@@ -7,8 +7,11 @@
 import tensorflow as tf
 from tensorflow.keras import Sequential, Model
 from tensorflow.keras.layers import Dense
-from tensorflow.keras.optimizers import Adam
 from tensorflow.keras.initializers import GlorotUniform
+try:
+    from tensorflow.keras.optimizers.legacy import Adam
+except:
+    from tensorflow.keras.optimizers import Adam
 
 from adapt.feature_based import ADDA
 

tests/test_base.py

@@ -9,7 +9,10 @@
 import tensorflow as tf
 from tensorflow.keras import Sequential, Model
 from tensorflow.keras.layers import Dense
-from tensorflow.keras.optimizers import Adam
+try:
+    from tensorflow.keras.optimizers.legacy import Adam
+except:
+    from tensorflow.keras.optimizers import Adam
 from sklearn.utils.estimator_checks import check_estimator
 from sklearn.base import clone
 

tests/test_ccsa.py

@@ -4,6 +4,10 @@
 from adapt.utils import make_classification_da
 from adapt.feature_based import CCSA
 from tensorflow.keras.initializers import GlorotUniform
+try:
+    from tensorflow.keras.optimizers.legacy import Adam
+except:
+    from tensorflow.keras.optimizers import Adam
 
 np.random.seed(0)
 tf.random.set_seed(0)
@@ -18,13 +22,13 @@
 
 def test_ccsa():
     ccsa = CCSA(task=task, loss="categorical_crossentropy",
-                optimizer="adam", metrics=["acc"], gamma=0.1, random_state=0)
+                optimizer=Adam(), metrics=["acc"], gamma=0.1, random_state=0)
     ccsa.fit(Xs, tf.one_hot(ys, 2).numpy(), Xt=Xt[ind],
              yt=tf.one_hot(yt, 2).numpy()[ind], epochs=100, verbose=0)
     assert np.mean(ccsa.predict(Xt).argmax(1) == yt) > 0.8
 
     ccsa = CCSA(task=task, loss="categorical_crossentropy",
-                optimizer="adam", metrics=["acc"], gamma=1., random_state=0)
+                optimizer=Adam(), metrics=["acc"], gamma=1., random_state=0)
     ccsa.fit(Xs, tf.one_hot(ys, 2).numpy(), Xt=Xt[ind],
              yt=tf.one_hot(yt, 2).numpy()[ind], epochs=100, verbose=0)
 

tests/test_cdan.py

@@ -6,7 +6,10 @@
 import tensorflow as tf
 from tensorflow.keras import Sequential, Model
 from tensorflow.keras.layers import Dense
-from tensorflow.keras.optimizers import Adam
+try:
+    from tensorflow.keras.optimizers.legacy import Adam
+except:
+    from tensorflow.keras.optimizers import Adam
 from tensorflow.keras.initializers import GlorotUniform
 
 from adapt.feature_based import CDAN

tests/test_dann.py

@@ -7,7 +7,10 @@
 import tensorflow as tf
 from tensorflow.keras import Sequential, Model
 from tensorflow.keras.layers import Dense
-from tensorflow.keras.optimizers import Adam
+try:
+    from tensorflow.keras.optimizers.legacy import Adam
+except:
+    from tensorflow.keras.optimizers import Adam
 
 from adapt.feature_based import DANN
 from adapt.utils import UpdateLambda