fixed denodogram + xgboost label encofding

barahona-research-group · Jun 4, 2024 · 83dec3b · 83dec3b
1 parent 8bfdf34
commit 83dec3b
Show file tree

Hide file tree

Showing 2 changed files with 5 additions and 2 deletions.
diff --git a/hcga/analysis.py b/hcga/analysis.py
@@ -14,7 +14,7 @@
 import shap
 from sklearn.metrics import accuracy_score, mean_absolute_error
 from sklearn.model_selection import RepeatedKFold, RepeatedStratifiedKFold, ShuffleSplit
-from sklearn.preprocessing import StandardScaler
+from sklearn.preprocessing import StandardScaler, LabelEncoder
 from tqdm import tqdm
 
 from hcga.io import load_fitted_model, save_fitted_model
@@ -684,7 +684,10 @@ def classify_pairwise(  # pylint: disable=too-many-locals
 
     top_features = {}
     for pair in tqdm(class_pairs):
+        L.info("Pairwise classification between classes %s and %s", str(pair[0]), str(pair[1]))
         features_pair = features.loc[(features.label == pair[0]) | (features.label == pair[1])]
+        le = LabelEncoder()
+        features_pair.label = le.fit_transform(features_pair.label)
         analysis_results = fit_model_kfold(
             features_pair,
             classifier,

diff --git a/hcga/plotting.py b/hcga/plotting.py
@@ -194,7 +194,7 @@ def _plot_dendrogram_shap(
     ax2 = plt.subplot(gs[1, 0])
 
     cor = np.abs(X_red.corr())
-    Z = linkage(cor.to_numpy(), "ward")
+    Z = linkage(np.nan_to_num(cor.to_numpy()), "ward")
     dn = dendrogram(Z, labels=X_red.columns, ax=ax1)
     ax1.xaxis.set_ticklabels([])
     ax1.set_ylabel("Euclidean Distance")