feat: model

README.md

@@ -9,4 +9,4 @@ app_file: app.py
 pinned: false
 ---
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+This repository contains an implementation of an App for a health insurance machine learning model.

models/HealthInsurance.py

@@ -0,0 +1,58 @@
+
+import pandas as pd
+
+
+class HealthInsurance():
+
+    def __init__(self, model, column_transformer, bins_annual_premium_type):
+        """model : the sklearn model already trainned.
+           colums_transformer : the column transformer with all transformations.
+           bins_annual_premium_type : bins to create annual_premium_type feature"""
+
+        self.model = model
+        self.transformer = column_transformer
+        self.bins_annual_premium_type = bins_annual_premium_type
+
+
+    def feature_engineering(self, df):
+        premium_categories = ['very_low', 'low', 'moderate', 'high', 'very_high']
+
+        df['vehicle_age'] = df['vehicle_age'].apply(self.get_vehicle_age)
+        df['annual_premium_type'] = pd.cut(x = df['annual_premium'], 
+                                            bins = self.bins_annual_premium_type,
+                                            labels = premium_categories)
+        return df
+
+    def get_vehicle_age(self, vehicle_age):
+
+        vehicle_labels = { 
+            '> 2 Years' : 'over_2_years',
+            '1-2 Year' : 'between_1_2_year',
+            '< 1 Year' : 'below_1_year'
+            }
+        
+        return vehicle_labels.get(vehicle_age)
+
+    def data_preparation(self, df):        
+        return self.transformer.transform(df)
+    
+    def predict(self, payload):
+        df = pd.read_json(payload, orient='records')
+
+        np_array = (df.pipe(self.feature_engineering)
+                         .pipe(self.data_preparation)
+                    )
+
+        df['score'] = self.model.predict_proba(np_array)[:, 1]
+
+        return df.to_json(orient='records')
+
+
+
+    
+
+
+
+
+
+

models/__init__.py

@@ -0,0 +1 @@
+from .HealthInsurance import HealthInsurance

parameters/bins_annual_premium_type.joblib

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:e42fdac17630e21c47134283f0690ec69c912e2b9dbd0fd68baaef6319d847ee
+size 231

parameters/column_transformer.joblib

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:ed9a96c0f7a377578494f6157cfc6ee49b27ac5c1b9ac4cff4b349297d3776cd
+size 22746

parameters/random_forrest.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:4d4ffc09d8834ea81a6006d495db91f41621adfff13847c7930f273d7e1ebcf4
+size 86534189

requirements.txt

@@ -0,0 +1,23 @@
+category-encoders==2.5.0
+click==8.1.3
+deprecation==2.1.0
+Flask==2.2.2
+itsdangerous==2.1.2
+Jinja2==3.1.2
+joblib==1.1.0
+MarkupSafe==2.1.1
+numpy==1.23.2
+packaging==21.3
+pandas==1.4.4
+patsy==0.5.2
+pyparsing==3.0.9
+python-dateutil==2.8.2
+pytz==2022.2.1
+scikit-learn==1.1.2
+scipy==1.9.1
+six==1.16.0
+sklearn==0.0
+statsmodels==0.13.2
+threadpoolctl==3.1.0
+watchdog==2.1.9
+Werkzeug==2.2.2