How to Build a Real‑Time People Flow Heatmap with Python, Pandas & Folium

While browsing Tencent's big‑data map, I noticed it visualizes user locations from services like WeChat and QQ. Inspired, I created a Python script to generate custom regional foot‑traffic heatmaps.

1. Data Acquisition

Tencent provides a data API for commercial use, but we can capture the endpoint by monitoring network requests on https://xingyun.map.qq.com/ . Each request sends four POST parameters (count, rank, etc.). The response contains a locs field where the first two numbers represent latitude (e.g., 3295) and longitude (e.g., 11590) multiplied by 100, and the third number indicates the number of people at that point.

Below is the code to fetch the data:

import requests
import json
header = {'User-Agent': 'Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:82.0) Gecko/20100101 Firefox/82.0'}
url = 'https://xingyun.map.qq.com/api/getXingyunPoints'
for i in range(1, 5):
    payload = {'count': i, 'rank': 0}
    response = requests.post(url, data=json.dumps(payload))
    datas = json.loads(response.text)['locs']
    datas = datas.split(',')
    datas = [int(x) for x in datas[:-1]]
    all_data = []
    a = []
    for n, data in enumerate(datas):
        a.append(data)
        all_data.append(a)
        if (n + 1) % 3 == 0:
            a = []
all_data = [[i[0]/100, i[1]/100, i[2]] for i in all_data]

Convert the list to a pandas DataFrame:

import pandas as pd
lat = [float(i[0]) for i in all_data]
lon = [i[1] for i in all_data]
weight = [i[2] for i in all_data]
dataframe = pd.DataFrame({'纬度': lat, '经度': lon, '人数': weight})

Remove duplicate entries:

dataframe = dataframe.drop_duplicates(keep='first')

Pandas tip

DataFrame.drop_duplicates(subset=None, keep='first', inplace=False)

removes duplicate rows; keep='first' retains the first occurrence.

Select a specific region

data1 = dataframe[(dataframe.纬度.between(39.26, 41.03)) & (dataframe.经度.between(115.25, 117.30))]

2. Draw the heatmap with Folium

import folium
from folium.plugins import HeatMap
map_data = data1[['纬度', '经度', '人数']].values.tolist()
hmap = folium.Map(location=[data1['纬度'].mean(), data1['经度'].mean()], control_scale=True, zoom_start=13)
hmap.add_child(HeatMap(map_data, radius=5, gradient={.1: 'blue', .3: 'lime', .5: 'yellow', .7: 'red'}))

The resulting map visualizes the estimated foot‑traffic density for the chosen area.