Recipe Suggestions From Connections

Problem

A cooking community lets members connect and save recipes.

DataFrame `connections` (from `connections.csv`):

```text
+----------+------+
| Column | Type |
+----------+------+
| member_a | int |
| member_b | int |
+----------+------+
(member_a, member_b) is the primary key. Each row means member_a and member_b are connected. Connections are mutual but each pair is stored only once (in one direction).
```

DataFrame `saves` (from `saves.csv`):

```text
+-----------+------+
| Column | Type |
+-----------+------+
| member_id | int |
| recipe_id | int |
+-----------+------+
(member_id, recipe_id) is the primary key. Each row means a member saved a recipe.
```

The app suggests a recipe to a member when at least one of that member's connections has saved it **and** the member has not already saved it themselves. For every member, report how many of their connections saved each suggested recipe.

Return a DataFrame with columns `member_id`, `recipe_id`, and `connection_saves` (the count of connections who saved it), in any order.

Input data

Example rows — the live problem includes the full dataset.

Expected output

Your answer should return 16 rows with the columns member_id, recipe_id, connection_saves.

Starter code (Pandas (Python))

import pandas as pd

def recipe_suggestions_from_connections(connections, saves) -> pd.DataFrame:
    # Your code here
    return connections

Related Pandas questions

• Always Mid-Pack BakersHard
• Annual Energy Output Growth by Panel ModelHard
• Attendees By TrackHard
• Auto-Tagging Support TicketsHard
• Bake-Off Division ChampionsHard
• Branch Energy Spend vs Chain AverageHard