UP my solution

pandas_questions.py

@@ -1,23 +1,24 @@
 """Plotting referendum results in pandas.
 
-In short, we want to make beautiful map to report results of a referendum. In
+In short, we want to make a 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
 import geopandas as gpd
 import matplotlib.pyplot as plt
 
 
 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', sep=';')
+    regions = pd.read_csv('data/regions.csv')
+    departments = pd.read_csv('data/departments.csv')
 
     return referendum, regions, departments
 
@@ -28,18 +29,39 @@ def merge_regions_and_departments(regions, departments):
     The columns in the final DataFrame should be:
     ['code_reg', 'name_reg', 'code_dep', 'name_dep']
     """
+    merged_df = pd.merge(
+        departments, regions,
+        left_on='region_code', right_on='code',
+        suffixes=('_dep', '_reg')
+    )
+    regions_and_departments = merged_df[
+        ['code_reg', 'name_reg', 'code_dep', 'name_dep']
+    ]
+    regions_and_departments.columns = [
+        'code_reg', 'name_reg', 'code_dep', 'name_dep'
+        ]
 
-    return pd.DataFrame({})
+    return regions_and_departments
 
 
 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.
     """
+    regions_and_departments['code_dep'] = regions_and_departments[
+        'code_dep'].apply(lambda x: str(x).lstrip('0'))
+    indices_to_drop = referendum[
+        referendum['Department code'].str.startswith('Z')
+    ].index
+    filtered_referendum = referendum.drop(indices_to_drop)
+    referendum_and_areas = pd.merge(
+        regions_and_departments, filtered_referendum,
+        right_on='Department code', left_on='code_dep', how='right'
+    )
 
-    return pd.DataFrame({})
+    return referendum_and_areas
 
 
 def compute_referendum_result_by_regions(referendum_and_areas):
@@ -48,8 +70,20 @@ 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().reset_index()
+
+    return grouped.set_index('code_reg')[
+        [
+            'name_reg',
+            'Registered',
+            'Abstentions',
+            'Null',
+            'Choice A',
+            'Choice B'
+        ]
+    ]
 
 
 def plot_referendum_map(referendum_result_by_regions):
@@ -61,8 +95,19 @@ 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.
     """
+    regions_geo = gpd.read_file('data/regions.geojson')
+    merged = pd.merge(
+        regions_geo, referendum_result_by_regions,
+        left_on='code', right_on='code_reg'
+    )
+    merged['ratio'] = merged['Choice A'] / (
+        merged['Choice A'] + merged['Choice B']
+        )
+    merged.plot(
+        column='ratio', legend=True, cmap='coolwarm'
+    )
 
-    return gpd.GeoDataFrame({})
+    return merged
 
 
 if __name__ == "__main__":