Merge pull request #5 from SandroMartens/dev

Dev
SandroMartens · Mar 22, 2024 · acfc6d7 · acfc6d7
2 parents b65d424 + 8600b0b
commit acfc6d7
Show file tree

Hide file tree

Showing 12 changed files with 2,173 additions and 360 deletions.
diff --git a/dbgsom/BaseSom.py b/dbgsom/BaseSom.py
diff --git a/dbgsom/SomClassifier.py b/dbgsom/SomClassifier.py
@@ -12,6 +12,7 @@
     check_is_fitted,
     check_X_y,
 )
+
 from .BaseSom import BaseSom
 
 
@@ -126,8 +127,9 @@ def _check_input_data(
         X, y = check_X_y(X=X, y=y, ensure_min_samples=4, dtype=[np.float64, np.float32])
         return X, y
 
-    def _label_prototypes(self, X, y) -> None:
-        winners = self._get_winning_neurons(X, n_bmu=1)
+    def _label_prototypes(self, X: npt.ArrayLike, y=npt.ArrayLike) -> None:
+        """This method assigns labels to the prototypes based on the input data."""
+        _, winners = self._get_winning_neurons(X, n_bmu=1)
         for winner_index, neuron in enumerate(self.neurons_):
             labels = y[winners == winner_index]
             # dead neuron
@@ -150,8 +152,9 @@ def _label_prototypes(self, X, y) -> None:
                 )
 
     def _fit(self, X: npt.ArrayLike, y: None | npt.ArrayLike = None):
-        classes, y = np.unique(y, return_inverse=True)
-        self.classes_ = classes
+        pass
+        # classes, y = np.unique(y, return_inverse=True)
+        # self.classes_ = classes
 
     def predict(self, X: npt.ArrayLike) -> np.ndarray:
         """Predict class labels for samples in X.
@@ -190,7 +193,7 @@ def predict_proba(self, X: npt.ArrayLike) -> np.ndarray:
         check_is_fitted(self)
         X = check_array(X)
         if self.vertical_growth:
-            winners = self._get_winning_neurons(X, n_bmu=1)
+            _, winners = self._get_winning_neurons(X, n_bmu=1)
             probabilities_rows = []
             for sample, winner in zip(X, winners):
                 node = self.neurons_[winner]
@@ -206,7 +209,6 @@ def predict_proba(self, X: npt.ArrayLike) -> np.ndarray:
             sample_probabilities = np.array(probabilities_rows)
 
         else:
-            # pass
             X_transformed = self.transform(X)
             node_probabilities = self._extract_values_from_graph("probabilities")
             # Sample Probabilities do not sum to 1

diff --git a/dbgsom/SomVQ.py b/dbgsom/SomVQ.py
@@ -3,7 +3,12 @@
 
 import numpy as np
 import numpy.typing as npt
-from sklearn.base import ClusterMixin, TransformerMixin, check_array, check_is_fitted
+from sklearn.base import (
+    ClusterMixin,
+    TransformerMixin,
+    check_array,
+    check_is_fitted,
+)
 
 from .BaseSom import BaseSom
 
@@ -115,7 +120,8 @@ class SomVQ(BaseSom, ClusterMixin, TransformerMixin):
 
     def _check_input_data(self, X: npt.ArrayLike, y=None) -> tuple[npt.NDArray, None]:
         X = check_array(array=X, ensure_min_samples=4, dtype=[np.float64, np.float32])
-        return X, y
+        # throw away any y
+        return X, None
 
     def _label_prototypes(self, X: npt.ArrayLike, y=None) -> None:
         for i, neuron in enumerate(self.som_):
@@ -137,6 +143,10 @@ def predict(self, X: npt.ArrayLike) -> np.ndarray:
         """
         check_is_fitted(self)
         X = check_array(X)
-        labels = self._get_winning_neurons(X, n_bmu=1)
+        _, labels = self._get_winning_neurons(X, n_bmu=1)
 
         return labels
+
+    def _fit(self, X: npt.NDArray):
+
+        self.labels_ = self.predict(X)
diff --git a/dbgsom/dbgsom.test b/dbgsom/dbgsom.test
diff --git a/examples/2d_example.ipynb b/examples/2d_example.ipynb
diff --git a/examples/2d_example.png b/examples/2d_example.png
diff --git a/examples/chain_link.ipynb b/examples/chain_link.ipynb
diff --git a/examples/darknet.ipynb b/examples/darknet.ipynb
diff --git a/examples/digits.ipynb b/examples/digits.ipynb
@@ -4,21 +4,34 @@
    "cell_type": "code",
    "execution_count": 1,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "ename": "ModuleNotFoundError",
+     "evalue": "No module named 'dbgsom'",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[1;31mModuleNotFoundError\u001b[0m                       Traceback (most recent call last)",
+      "Cell \u001b[1;32mIn[1], line 9\u001b[0m\n\u001b[0;32m      6\u001b[0m \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01msklearn\u001b[39;00m\u001b[38;5;21;01m.\u001b[39;00m\u001b[38;5;21;01mpipeline\u001b[39;00m \u001b[38;5;28;01mimport\u001b[39;00m Pipeline\n\u001b[0;32m      7\u001b[0m \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01msklearn\u001b[39;00m\u001b[38;5;21;01m.\u001b[39;00m\u001b[38;5;21;01mpreprocessing\u001b[39;00m \u001b[38;5;28;01mimport\u001b[39;00m StandardScaler\n\u001b[1;32m----> 9\u001b[0m \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01mdbgsom\u001b[39;00m\u001b[38;5;21;01m.\u001b[39;00m\u001b[38;5;21;01mdbgsom_\u001b[39;00m \u001b[38;5;28;01mimport\u001b[39;00m DBGSOM\n",
+      "\u001b[1;31mModuleNotFoundError\u001b[0m: No module named 'dbgsom'"
+     ]
+    }
+   ],
    "source": [
-    "import seaborn.objects as so\n",
-    "import seaborn as sns\n",
-    "import numpy as np\n",
     "import matplotlib.pyplot as plt\n",
-    "from dbgsom.dbgsom_ import DBGSOM\n",
-    "from sklearn.preprocessing import StandardScaler\n",
+    "import numpy as np\n",
+    "import seaborn as sns\n",
+    "import seaborn.objects as so\n",
+    "from sklearn.datasets import load_digits\n",
     "from sklearn.pipeline import Pipeline\n",
-    "from sklearn.datasets import load_digits"
+    "from sklearn.preprocessing import StandardScaler\n",
+    "\n",
+    "from dbgsom.dbgsom_ import DBGSOM"
    ]
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": null,
    "metadata": {},
    "outputs": [
     {
@@ -70,7 +83,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": null,
    "metadata": {},
    "outputs": [
     {
@@ -114,7 +127,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": null,
    "metadata": {},
    "outputs": [
     {
@@ -134,7 +147,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": null,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -144,7 +157,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": null,
    "metadata": {},
    "outputs": [
     {
@@ -174,7 +187,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": null,
    "metadata": {},
    "outputs": [
     {
@@ -206,7 +219,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": null,
    "metadata": {},
    "outputs": [
     {
@@ -253,7 +266,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.10.13"
+   "version": "3.12.1"
   },
   "orig_nbformat": 4
  },