Ring Division: Investment Analysis of Beijing Real Estate
In this project, our team conducted an in-depth data analysis and investment strategy study based on 400,000 Beijing real estate transaction records. We aimed to identify key factors influencing housing prices and provide ring-based investment recommendations to guide corporate investment decisions in Beijing's real estate market.
-
Used Excel for data cleaning (format conversion, outlier removal, deduplication).
-
Augmented data by web scraping with Python, converting geographic coordinates to WGS 84, and storing results in Excel.
-
Loaded global map data via GeoPandas, filtered for Beijing, and rendered district-level housing distributions using Shapefiles.
-
Calculated average followers per district and generated interactive visualizations in Tableau.
-
Innovatively calculated ring areas (3rd Ring, 4th Ring, etc.) using C++, then mapped them with Python/Pandas.
-
Ring-based segmentation: Analyzed price drivers in 4 zones (Inside 3rd Ring, 3rd–4th Ring, 4th–5th Ring, Beyond 5th Ring).
-
Regression analysis: Evaluated impacts of factors like construction year, transaction date, area, subway proximity, elevator availability, and "Five-Year Property Policy".
-
Key finding: Followers count had low variance and was not a reliable investment indicator; prioritized location-centric metrics.
-
Delivered interactive Tableau dashboards with heatmaps, trend analyses, and ring-wise comparisons.
-
Proposed targeted investment strategies for each ring zone, supported by statistical evidence.
-
Provided actionable insights for corporate investment portfolios, optimizing risk/return trade-offs.
将团队设想为数据分析师，从近40万条北京房产交易数据中挖掘值得探讨的问题和现象。依次列出一级问题、二级问题和三级问题，从这些问题出发，找出投资北京房地产的正确方式，并尝试说服投资北京房地产的上市公司按照我们的建议进行投资。（具体项目要求见根目录下的Project.pdf）
每条房产记录包含众多属性，包括但不限于：房屋建造时间、交易时间、经纬度坐标、每平米价格、总价格、房间数量、总面积、是否临近地铁、是否有电梯、楼层、是否符合五年产权政策、关注人数等。
为帮助更多人找到合适的房子，我们尝试从关注人数（followers）入手。然而，分析发现关注人数的极差和方差均较小，这不利于房地产投资分析。我们推测，人们往往会关注自己能买得起的房子以及渴望购买的房子，因此，关注人数并不能很好地反映市场情况。
讨论不同环区域的价格影响因素。由于团队中有成员熟悉北京，意识到不同环区域之间存在显著差异，因此将数据分为3环内、3到4环、4到5环和5环外四个区域进行分析。这种分区符合人们的购房思路。
为了更清晰地展示各个影响因素在不同环区域中的影响程度，我们的分工是每个人分析几个属性作为自变量，研究这些自变量对房价的影响，而不是按环区域进行分工。
最终，经过结论的整合与汇总，我们为不同环区域的投资分别提出了相应建议，并针对全市的共性特点提出了针对北京房地产投资的总体建议。（详细的结论和成果见submit文件夹和根目录下的Poster.png）
此处仅展示最终数据大屏的结果：
