Upload my_module.py

my_module.py

@@ -0,0 +1,44 @@
+from sklearn.model_selection import GridSearchCV
+from sklearn.metrics import confusion_matrix, precision_score, recall_score, accuracy_score, f1_score
+
+def my_evaluation(y, pred):
+  # 正解と予測値を入れると自動で評価してくれます。
+  print('混同行列:')
+  print(confusion_matrix(y, pred))
+  accuracy = accuracy_score(y, pred)
+  precision = precision_score(y, pred)
+  recall = recall_score(y, pred)
+  f1 = f1_score(y, pred)
+
+  print("正解率: %.3f" % accuracy)
+  print("精度: %.3f" % precision)
+  print("再現率: %.3f" % recall)
+  print("F1スコア: %.3f" % f1)
+
+
+# 万能関数
+def my_model(X_train, X_test, y_train, y_test, model, name=''):
+  # モデル訓練
+  model.fit(X_train, y_train)
+
+  pred = model.predict(X_test)# 予測
+  # モデル評価
+  print(f'モデル名:{name}')
+  my_evaluation(y_test, pred)
+  return model
+
+def my_best_model(X_train, X_test, y_train, y_test, model,params, name=''):
+  # モデル訓練
+  # 基礎モデル、パラメーター集合
+  models = GridSearchCV(model, params)
+  models.fit(X_train, y_train)
+
+  best_model = models.best_estimator_
+  pred = best_model.predict(X_test)# 予測
+  # モデル評価
+  print(f'モデル名:{name}')
+  my_evaluation(y_test, pred)
+  return best_model
+
+
+