renaming, issue #294 (#299)

Addressed issue #294 about changing every occurence of "concept_drift" to "conditional_shift"
Also changed "concept drift" to "conditional shift" in docstrings

Author: mbarneche

conftest.py

@@ -26,7 +26,7 @@ def da_reg_dataset():
     return make_shifted_datasets(
         n_samples_source=20,
         n_samples_target=21,
-        shift="concept_drift",
+        shift="conditional_shift",
         mean=0.5,
         noise=0.3,
         label="regression",
@@ -40,7 +40,7 @@ def da_reg_datasets():
     da_reg_dataset_1 = make_shifted_datasets(
         n_samples_source=5,
         n_samples_target=10,
-        shift="concept_drift",
+        shift="conditional_shift",
         mean=0.5,
         noise=0.3,
         label="regression",
@@ -51,7 +51,7 @@ def da_reg_datasets():
     da_reg_dataset_2 = make_shifted_datasets(
         n_samples_source=10,
         n_samples_target=5,
-        shift="concept_drift",
+        shift="conditional_shift",
         mean=0.5,
         noise=0.3,
         label="regression",
@@ -65,7 +65,7 @@ def da_multiclass_dataset():
     return make_shifted_datasets(
         n_samples_source=20,
         n_samples_target=21,
-        shift="concept_drift",
+        shift="conditional_shift",
         noise=0.1,
         label="multiclass",
         random_state=42,
@@ -78,7 +78,7 @@ def da_binary_dataset():
     return make_shifted_datasets(
         n_samples_source=20,
         n_samples_target=21,
-        shift="concept_drift",
+        shift="conditional_shift",
         noise=0.1,
         label="binary",
         random_state=42,

examples/datasets/plot_shifted_dataset.py

@@ -5,7 +5,7 @@
 This illustrates the :func:`~skada.datasets.make_shifted_dataset`
 dataset generator. Each method consists of generating source data
 and shifted target data. We illustrate here:
-covariate shift, target shift, concept drift, and sample bias.
+covariate shift, target shift, conditional shift, and sample bias.
 See detailed description of each shift in [1]_.
 
 .. [1] Moreno-Torres, J. G., Raeder, T., Alaiz-Rodriguez,
@@ -27,7 +27,7 @@ def plot_shifted_dataset(shift, random_state=42):
     """Plot source and shifted target data for a given type of shift.
 
     The possible shifts are 'covariate_shift', 'target_shift',
-    'concept_drift', or 'subspace'.
+    'conditional_shift', or 'subspace'.
 
     We use here the same random seed for multiple calls to
     ensure same distributions.
@@ -88,7 +88,7 @@ def plot_shifted_dataset(shift, random_state=42):
 for shift in [
     "covariate_shift",
     "target_shift",
-    "concept_drift",
+    "conditional_shift",
     "subspace",
 ]:
     plot_shifted_dataset(shift)

examples/datasets/plot_shifted_dataset_regression.py

@@ -5,7 +5,7 @@
 This illustrates the :func:`~skada.datasets.make_shifted_dataset`
 dataset generator. Each method consists of generating source data
 and shifted target data. We illustrate here:
-covariate shift, target shift, concept drift, and sample bias.
+covariate shift, target shift, conditional shift, and sample bias.
 See detailed description of each shift in [1]_.
 
 .. [1] Moreno-Torres, J. G., Raeder, T., Alaiz-Rodriguez,
@@ -32,7 +32,7 @@ def plot_shifted_dataset(shift, random_state=42):
     """Plot source and shifted target data for a given type of shift.
 
     The possible shifts are 'covariate_shift', 'target_shift',
-    'concept_drift' or 'subspace'.
+    'conditional_shift' or 'subspace'.
 
     We use here the same random seed for multiple calls to
     ensure same distributions.
@@ -86,5 +86,5 @@ def plot_shifted_dataset(shift, random_state=42):
 
 # %% Visualize shifted datasets
 
-for shift in ["covariate_shift", "target_shift", "concept_drift", "subspace"]:
+for shift in ["covariate_shift", "target_shift", "conditional_shift", "subspace"]:
     plot_shifted_dataset(shift)

examples/methods/plot_jdot_da.py

@@ -3,7 +3,7 @@
 ======================================
 
 This example shows how to use the JDOTRegressor [10] to learn a regression model
-from source to target domain on a simple concept drift 2D example. We use a
+from source to target domain on a simple conditional shift 2D example. We use a
 simple Kernel Ridge Regression (KRR) as base estimator.
 
 We compare the performance of the KRR on the source and target domain, and the
@@ -33,15 +33,15 @@
 from skada.datasets import make_shifted_datasets
 
 # %%
-# Generate concept drift regression dataset and plot it
+# Generate conditional shift regression dataset and plot it
 # -----------------------------------------------------
 #
-# We generate a simple 2D concept drift dataset.
+# We generate a simple 2D conditional shift dataset.
 
 X, y, sample_domain = make_shifted_datasets(
     n_samples_source=20,
     n_samples_target=20,
-    shift="concept_drift",
+    shift="conditional_shift",
     noise=0.3,
     label="regression",
     random_state=42,
@@ -170,15 +170,15 @@
 
 
 # %%
-# Generate concept drift classification dataset and plot it
+# Generate conditional shift classification dataset and plot it
 # -----------------------------------------------------
 #
-# We generate a simple 2D concept drift dataset.
+# We generate a simple 2D conditional shift dataset.
 
 X, y, sample_domain = make_shifted_datasets(
     n_samples_source=20,
     n_samples_target=20,
-    shift="concept_drift",
+    shift="conditional_shift",
     noise=0.2,
     label="multiclass",
     random_state=42,

examples/methods/plot_label_prop_da.py

@@ -7,7 +7,7 @@
 the target domain using the OT plan. This was proposed originally in [28]_ for
 semi-supervised learning but can be used for DA.
 
-We illustrate the method on a simple regression and classification concept drift
+We illustrate the method on a simple regression and classification conditional shift
 dataset. We train a simple Kernel Ridge Regression (KRR) and Logistic Regression
 on the source domain and evaluate their performance on the source and target
 domain. We then train the same models with the label propagation method and
@@ -42,15 +42,15 @@
 from skada.datasets import make_shifted_datasets
 
 # %%
-# Generate concept drift regression dataset and plot it
+# Generate conditional shift regression dataset and plot it
 # -----------------------------------------------------
 #
-# We generate a simple 2D concept drift dataset.
+# We generate a simple 2D conditional shift dataset.
 
 X, y, sample_domain = make_shifted_datasets(
     n_samples_source=20,
     n_samples_target=20,
-    shift="concept_drift",
+    shift="conditional_shift",
     noise=0.3,
     label="regression",
     random_state=42,
@@ -177,15 +177,15 @@
 
 
 # %%
-# Generate concept drift classification dataset and plot it
+# Generate conditional shift classification dataset and plot it
 # -----------------------------------------------------
 #
-# We generate a simple 2D concept drift dataset.
+# We generate a simple 2D conditional shift dataset.
 
 X, y, sample_domain = make_shifted_datasets(
     n_samples_source=20,
     n_samples_target=20,
-    shift="concept_drift",
+    shift="conditional_shift",
     noise=0.2,
     label="multiclass",
     random_state=42,

examples/methods/plot_monge_alignment_da.py

@@ -24,15 +24,15 @@
 from skada.datasets import make_shifted_datasets
 
 # %%
-# Generate concept drift classification dataset and plot it
+# Generate conditional shift classification dataset and plot it
 # -----------------------------------------------------
 #
-# We generate a simple 2D concept drift dataset.
+# We generate a simple 2D conditional shift dataset.
 
 X, y, sample_domain = make_shifted_datasets(
     n_samples_source=20,
     n_samples_target=20,
-    shift="concept_drift",
+    shift="conditional_shift",
     noise=0.2,
     label="multiclass",
     random_state=42,
@@ -136,7 +136,7 @@ def get_multidomain_data(
         n_samples_source=n_samples_source,
         n_samples_target=n_samples_target,
         noise=noise,
-        shift="concept_drift",
+        shift="conditional_shift",
         label="multiclass",
         random_state=random_state,
     )
@@ -145,7 +145,7 @@ def get_multidomain_data(
             n_samples_source=n_samples_source,
             n_samples_target=n_samples_target,
             noise=noise,
-            shift="concept_drift",
+            shift="conditional_shift",
             label="multiclass",
             random_state=random_state + ns,
             mean=np.random.randn(2),
@@ -161,7 +161,7 @@ def get_multidomain_data(
             n_samples_source=n_samples_source,
             n_samples_target=n_samples_target,
             noise=noise,
-            shift="concept_drift",
+            shift="conditional_shift",
             label="multiclass",
             random_state=random_state + nt + 42,
             mean=np.random.randn(2),

examples/methods/plot_optimal_transport_da.py

@@ -31,13 +31,13 @@
 from skada.datasets import make_shifted_datasets
 
 # %%
-# Generate concept drift dataset
+# Generate conditional shift dataset
 # ------------------------------
 n_samples = 20
 X, y, sample_domain = make_shifted_datasets(
     n_samples_source=n_samples,
     n_samples_target=n_samples + 1,
-    shift="concept_drift",
+    shift="conditional_shift",
     noise=0.1,
     random_state=42,
 )

examples/plot_how_to_use_skada.py

@@ -50,7 +50,7 @@
 
 # Get DA dataset
 X, y, sample_domain = make_shifted_datasets(
-    20, 20, shift="concept_drift", random_state=42
+    20, 20, shift="conditional_shift", random_state=42
 )
 
 # split source and target for visualization
@@ -137,7 +137,7 @@
 print("Accuracy on target:", pipe.score(Xt, yt))
 
 # create a DA pipeline with GaussianReweight adapter (does not work well on
-# concept drift).
+# conditional shift).
 pipe = make_da_pipeline(
     StandardScaler(),
     GaussianReweightAdapter(),

examples/plot_method_comparison.py

@@ -114,7 +114,7 @@
     make_shifted_datasets(
         n_samples_source=20,
         n_samples_target=20,
-        shift="concept_drift",
+        shift="conditional_shift",
         label="binary",
         noise=0.4,
         random_state=RANDOM_SEED,

examples/validation/plot_cross_val_score_for_da.py

@@ -23,7 +23,7 @@
 dataset = make_shifted_datasets(
     n_samples_source=30,
     n_samples_target=20,
-    shift="concept_drift",
+    shift="conditional_shift",
     label="binary",
     noise=0.4,
     random_state=RANDOM_SEED,