WIP - Rough scaffolding for feature extractor and classifier.

pyproject.toml

@@ -17,6 +17,8 @@ classifiers = [
 dynamic = ["version"]
 requires-python = ">=3.9"
 dependencies = [
+    "resspect",
+    "scikit_learn",
 ]
 
 [project.urls]

src/laiss_resspect_classifier/laiss_classifier.py

@@ -0,0 +1,12 @@
+from sklearn.ensemble import RandomForestClassifier
+
+from resspect.classifiers import ResspectClassifier
+
+class LaissRandomForest(ResspectClassifier):
+    """LAISS-specific version of the sklearn RandomForestClassifier."""
+
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+
+        self.n_estimators = self.kwargs.pop('n_estimators', 100)
+        self.classifier = RandomForestClassifier(n_estimators=self.n_estimators, **self.kwargs)

src/laiss_resspect_classifier/laiss_feature_extractor.py

@@ -0,0 +1,13 @@
+import numpy as np
+
+from resspect.feature_extractors import LightCurve
+
+class LaissFeatureExtractor(LightCurve):
+    def __init__(self, **kwargs):
+        super().__init__(**kwargs)
+
+    def fit(self, band:str = None) -> np.ndarray:
+        pass
+
+    def fit_all(self) -> np.ndarray:
+        pass