Add Python lab solution

solution.py

@@ -0,0 +1,82 @@
+import pandas as pd
+from sqlalchemy import create_engine
+
+# ---------------------------
+# 1) Connect Python to Sakila
+# ---------------------------
+# Update these credentials
+# Example for MySQL:
+# mysql+mysqlconnector://USER:PASSWORD@HOST:PORT/sakila
+engine = create_engine("mysql+mysqlconnector://root:YOUR_PASSWORD@localhost:3306/sakila")
+
+
+# ----------------------------------------------
+# 2) rentals_month(engine, month, year) -> df
+# ----------------------------------------------
+def rentals_month(engine, month: int, year: int) -> pd.DataFrame:
+    """
+    Retrieve rental rows for a given month/year from Sakila rental table.
+    Returns a pandas DataFrame.
+    """
+    sql = """
+        SELECT rental_id, rental_date, customer_id, inventory_id, return_date, staff_id
+        FROM rental
+        WHERE YEAR(rental_date) = :year
+          AND MONTH(rental_date) = :month
+    """
+    return pd.read_sql(sql, engine, params={"year": year, "month": month})
+
+
+# -------------------------------------------------------------------
+# 3) rental_count_month(df, month, year) -> df with rentals_MM_YYYY
+# -------------------------------------------------------------------
+def rental_count_month(rentals_df: pd.DataFrame, month: int, year: int) -> pd.DataFrame:
+    """
+    Takes the rentals_month output and returns rentals per customer_id for that month/year.
+    Column name format: rentals_MM_YYYY (MM is zero-padded).
+    """
+    col_name = f"rentals_{month:02d}_{year}"
+
+    # groupby customer_id and count rentals
+    counts = (
+        rentals_df.groupby("customer_id")["rental_id"]
+        .count()
+        .reset_index(name=col_name)
+    )
+    return counts
+
+
+# -------------------------------------------------------------------
+# 4) compare_rentals(df1, df2) -> combined df + difference column
+# -------------------------------------------------------------------
+def compare_rentals(df_month1: pd.DataFrame, df_month2: pd.DataFrame) -> pd.DataFrame:
+    """
+    Takes two customer rental-count dataframes (one per month) and returns:
+    customer_id, rentals_MM_YYYY (month1), rentals_MM_YYYY (month2), difference
+
+    difference = month2 - month1 (so positive means increased in month2).
+    Keeps only customers present in BOTH dfs (active in both months).
+    """
+    # detect the rentals column names (besides customer_id)
+    rentals_col_1 = [c for c in df_month1.columns if c != "customer_id"][0]
+    rentals_col_2 = [c for c in df_month2.columns if c != "customer_id"][0]
+
+    merged = df_month1.merge(df_month2, on="customer_id", how="inner")
+
+    merged["difference"] = merged[rentals_col_2] - merged[rentals_col_1]
+    return merged.sort_values("difference", ascending=False).reset_index(drop=True)
+
+
+# ---------------------------
+# Example: May vs June (2005)
+# ---------------------------
+may_rentals = rentals_month(engine, 5, 2005)
+june_rentals = rentals_month(engine, 6, 2005)
+
+may_counts = rental_count_month(may_rentals, 5, 2005)
+june_counts = rental_count_month(june_rentals, 6, 2005)
+
+comparison = compare_rentals(may_counts, june_counts)
+
+print(comparison.head(20))          # top 20 biggest increases
+print("Customers active in both:", len(comparison))