Analyzing the Distribution and Competition Between Starbucks and Luckin Coffee Using Python Data Visualization

In this tutorial we use Python and the Shapely library to fetch, process, and visualize the 2022 nationwide store data of Starbucks and Luckin Coffee, comparing their quantity, regional distribution, and spatial relationship.

The datasets contain store name, latitude, longitude, and city, with a noted ~20% count error. Data are accessed via API calls, as shown in the first code snippet.

import requests<br/>headers = {"x-token": "你的鉴权token"}<br/>response = requests.get("http://app.chafer.nexadata.cn/openapi/v1/sheet/sht22nId5uouP2/records?size=1&page=1", headers=headers)<br/>print(response.json())

After retrieving the data, we import necessary libraries and extract all pages of records for both brands.

# 导入相关库<br/>import pandas as pd<br/>import requests<br/>import time<br/>from shapely.geometry import Point<br/>from shapely.geometry.polygon import Polygon

We then define helper functions to create a 500 m radius circle around each Starbucks store, convert coordinates to points, and count how many Luckin stores fall inside each circle.

def circle(data, radius):<br/>    center_latitude = float(data['维度'])<br/>    center_longitude = float(data['经度'])<br/>    center_point = Point(center_longitude, center_latitude)<br/>    circle = center_point.buffer(radius/111300)<br/>    polygon = Polygon(circle.exterior)<br/>    return polygon<br/><br/>def point(data):<br/>    center_latitude = float(data['维度'])<br/>    center_longitude = float(data['经度'])<br/>    return Point(center_longitude, center_latitude)<br/><br/>def is_inside(data):<br/>    polygon = data['Polygon']<br/>    n = 0<br/>    luckin_city = luckin[luckin['城市']==data['城市']]<br/>    for pt in luckin_city['Point']:<br/>        if polygon.contains(pt):<br/>            n += 1<br/>    return n

Applying these functions yields a new column Luckin_numbers indicating the count of Luckin stores within 500 m of each Starbucks location.

Key findings include:

On average, each Starbucks store is surrounded by 0.6 Luckin stores within 500 m.

The maximum number of Luckin stores near a single Starbucks is 7.

City‑level analysis shows Linyi has the highest average of 1.8 Luckin stores per Starbucks.

Visualizations (heat maps and top‑20 city bar charts) illustrate that Starbucks is concentrated in coastal first‑tier cities, while Luckin has a more dispersed presence, even in many third‑ and fourth‑tier inland cities.

The study demonstrates how simple Python scripts and geospatial libraries can uncover competitive spatial patterns between coffee chains, suggesting strategic insights for market expansion.