diff --git a/cleaning.py b/cleaning.py
new file mode 100644
index 0000000..b7c66a4
--- /dev/null
+++ b/cleaning.py
@@ -0,0 +1,48 @@
+#!/usr/bin/env python3
+from pandas import DataFrame, to_numeric
+
+
+def display_info(df: DataFrame) -> None:
+    print(df.all())
+    print(df.info())
+    print("\nNombre de valeurs manquantes par colonne :")
+    print(df.isna().sum())
+
+
+def drop_empty_appellation(df: DataFrame) -> DataFrame:
+
+    return df.dropna(subset=["Appellation"])
+
+
+def mean_score(df: DataFrame, col: str) -> DataFrame:
+    """
+    Calcule la moyenne d'une colonne de score par appellation.
+        - Convertit les valeurs en numériques, en remplaçant les non-convertibles par NaN
+        - Calcule la moyenne par appellation
+        - Remplace les NaN résultants par 0 
+
+    """
+    tmp = df[["Appellation", col]].copy()
+
+    tmp[col] = to_numeric(tmp[col], errors="coerce")
+
+    # moyenne par appellation
+    means = tmp.groupby("Appellation", as_index=False)[col].mean()
+    
+    means[col] = means[col].fillna(0)
+    
+    means = means.rename(columns={col: f"mean_{col}"})
+    
+    return means
+
+
+def mean_robert(df: DataFrame) -> DataFrame:
+    return mean_score(df, "Robert")
+
+
+def mean_robinson(df: DataFrame) -> DataFrame:
+    return mean_score(df, "Robinson")
+
+
+def mean_suckling(df: DataFrame) -> DataFrame:
+    return mean_score(df, "Suckling")
\ No newline at end of file
diff --git a/main.py b/main.py
index 5747f74..a87f052 100755
--- a/main.py
+++ b/main.py
@@ -5,19 +5,53 @@ from os.path import normpath, join
 from sys import argv
 from pandas import read_csv, DataFrame
 
+from cleaning import (display_info,
+                      drop_empty_appellation,
+                      mean_robert,
+                      mean_robinson,
+                      mean_suckling)
+
+
+def load_csv(filename: str) -> DataFrame:
+    path: str = normpath(join(getcwd(), filename))
+    return read_csv(path)
+
+
+def save_csv(df: DataFrame, out_filename: str) -> None:
+    df.to_csv(out_filename, index=False)
+
+
 def main() -> None:
     if len(argv) != 2:
-        raise ValueError(f"{argv[0]} <filename.csv>")
+        raise ValueError(f"Usage: {argv[0]} <filename.csv>")
+
+    df = load_csv(argv[1])
+
+    print("=== Avant nettoyage ===")
+    display_info(df)
+
+    df = drop_empty_appellation(df)
+    save_csv(df, "donnee_clean.csv")
+
+    print("\n=== Après nettoyage d'appellations manquantes ===")
+    display_info(df)
+    
+    #la moyenne des notes des vins pour chaque appellation
+    robert_means = mean_robert(df)
+    save_csv(robert_means, "mean_robert_by_appellation.csv")
+    print("\n=== moyenne Robert par appellation ===")
+    print(robert_means.head(10))
+
+    robinson_means = mean_robinson(df)
+    save_csv(robinson_means, "mean_robinson_by_appellation.csv")
+    print("\n===: moyennes Robinson par appellation ===")
+    print(robinson_means.head(10))
+    
+    suckling_means = mean_suckling(df)
+    save_csv(suckling_means, "mean_suckling_by_appellation.csv")
+    print("\n===: moyennes Suckling par appellation ===")
+    print(suckling_means.head(10))
 
-    path: str = normpath(join(getcwd(), argv[1]))
-    db: DataFrame = read_csv(path)
-    print(db.all())
-    print(db.info())
-    print("\nnombre de valeurs manquantes pour chaque colonne :")
-    print(db.isna().sum())
-    db = db.dropna(subset=["Appellation"])
-    db.to_csv("donnee_clean.csv", index=False)
-    print(db.isna().sum())
 
 if __name__ == "__main__":
     try: