UP my solution

pandas_questions.py

@@ -1,23 +1,14 @@
-"""Plotting referendum results in pandas.
-
-In short, we want to make beautiful map to report results of a referendum. In
-some way, we would like to depict results with something similar to the maps
-that you can find here:
-https://github.com/x-datascience-datacamp/datacamp-assignment-pandas/blob/main/example_map.png
-
-To do that, you will load the data as pandas.DataFrame, merge the info and
-aggregate them by regions and finally plot them on a map using `geopandas`.
-"""
-import pandas as pd
+# -*- coding: utf-8 -*-
 import geopandas as gpd
 import matplotlib.pyplot as plt
+import pandas as pd
 
 
 def load_data():
-    """Load data from the CSV files referundum/regions/departments."""
-    referendum = pd.DataFrame({})
-    regions = pd.DataFrame({})
-    departments = pd.DataFrame({})
+    """Load data from the CSV files referendum/regions/departments."""
+    referendum = pd.read_csv("data/referendum.csv", delimiter=";")
+    regions = pd.read_csv("data/regions.csv")
+    departments = pd.read_csv("data/departments.csv")
 
     return referendum, regions, departments
 
@@ -28,18 +19,42 @@ def merge_regions_and_departments(regions, departments):
     The columns in the final DataFrame should be:
     ['code_reg', 'name_reg', 'code_dep', 'name_dep']
     """
+    regions = regions.rename(columns={"code": "code_reg", "name": "name_reg"})
+    departments = departments.rename(
+        columns={
+            "region_code": "code_reg",
+            "code": "code_dep",
+            "name": "name_dep",
+        }
+    )
 
-    return pd.DataFrame({})
+    merged = departments.merge(regions, on="code_reg", how="right")
+    return merged[["code_reg", "name_reg", "code_dep", "name_dep"]]
 
 
 def merge_referendum_and_areas(referendum, regions_and_departments):
     """Merge referendum and regions_and_departments in one DataFrame.
 
     You can drop the lines relative to DOM-TOM-COM departments, and the
-    french living abroad.
+    French living abroad.
     """
 
-    return pd.DataFrame({})
+    def standardize_department_code(code):
+        if code.isnumeric():
+            return code.zfill(2)  # Zero-pad numeric codes
+        return code
+
+    referendum["Department code"] = referendum["Department code"].apply(
+        standardize_department_code
+    )
+    # Merge and resolve column name conflicts
+    merged = referendum.merge(
+        regions_and_departments,
+        left_on="Department code",
+        right_on="code_dep",
+        how="inner",
+    )
+    return merged
 
 
 def compute_referendum_result_by_regions(referendum_and_areas):
@@ -48,8 +63,35 @@ def compute_referendum_result_by_regions(referendum_and_areas):
     The return DataFrame should be indexed by `code_reg` and have columns:
     ['name_reg', 'Registered', 'Abstentions', 'Null', 'Choice A', 'Choice B']
     """
-
-    return pd.DataFrame({})
+    grouped = (
+        referendum_and_areas.groupby(["code_reg", "name_reg"])
+        .sum(numeric_only=True)
+        .reset_index()
+    )
+    result = grouped[
+        [
+            "code_reg",
+            "name_reg",
+            "Registered",
+            "Abstentions",
+            "Null",
+            "Choice A",
+            "Choice B",
+        ]
+    ]
+
+    # Ensure all regions are included
+    all_regions = referendum_and_areas[
+        ["code_reg", "name_reg"]
+    ].drop_duplicates()
+
+    # Join with explicit suffixes to avoid overlap
+    result = (
+        all_regions.merge(result, how="inner", on=["code_reg", "name_reg"])
+        .fillna(0)
+        .set_index("code_reg")
+    )
+    return result
 
 
 def plot_referendum_map(referendum_result_by_regions):
@@ -61,16 +103,29 @@ def plot_referendum_map(referendum_result_by_regions):
       should display the rate of 'Choice A' over all expressed ballots.
     * Return a gpd.GeoDataFrame with a column 'ratio' containing the results.
     """
+    geo_data = gpd.read_file("data/regions.geojson")
+    geo_data = geo_data.rename(columns={"code": "code_reg"})
+
+    # Compute the ratio for 'Choice A'
+    referendum_result_by_regions["ratio"] = referendum_result_by_regions[
+        "Choice A"
+    ] / (
+        referendum_result_by_regions["Choice A"]
+        + referendum_result_by_regions["Choice B"]
+    )
 
-    return gpd.GeoDataFrame({})
+    merged = geo_data.merge(referendum_result_by_regions, on="code_reg")
 
+    # Plot
+    merged.plot(column="ratio", legend=True, cmap="coolwarm")
+    plt.title("Referendum Results: Choice A Ratio by Region")
+
+    return merged
 
-if __name__ == "__main__":
 
+if __name__ == "__main__":
     referendum, df_reg, df_dep = load_data()
-    regions_and_departments = merge_regions_and_departments(
-        df_reg, df_dep
-    )
+    regions_and_departments = merge_regions_and_departments(df_reg, df_dep)
     referendum_and_areas = merge_referendum_and_areas(
         referendum, regions_and_departments
     )