zect7/Russian_apartment_price

How to Use the Russian Apartment Price Dataset (Kvartis)

Dataset: main_data.csv
GitHub Repository: zect-project/top_datasets
Size: 4,000 real estate listings (2025–2026 data)
Target variable: real_price (price in Russian Rubles, ₽)

---

📋 Dataset Overview

This dataset contains detailed information about apartments for sale in five major Russian cities:

• Moscow
• Petersburg (St. Petersburg)
• Novosibirsk
• Yekaterinburg
• Kazan

It is perfect for:

• Price prediction (regression)
• Market analysis
• Feature importance studies
• Learning CatBoost / handling categorical + numerical features
• Russian real-estate research

---

🏠 Columns Description

Column	Type	Description	Example
city	categorical	City where the apartment is located	Moscow
rooms	integer	Number of rooms	2, 3, 5
m2	integer	Total living area in square meters	65, 111
repair	categorical	Repair condition	not_needs, needs, absolutely_needs
floor	integer	Floor of the apartment	8, 45
all_floor	integer	Total number of floors in the building	20, 60
real_price	integer	Target — actual sale price in RUB	10189609

Repair categories meaning:

• not_needs — excellent condition, ready to move in
• needs — requires cosmetic repairs
• absolutely_needs — major renovation required

---

🚀 Quick Start

1. Load the data

import pandas as pd

df = pd.read_csv("main_data.csv")
print(df.shape)           # (4000, 7)
df.head()

Basic exploration

import matplotlib.pyplot as plt
import seaborn as sns

# Price distribution by city
sns.boxplot(data=df, x='city', y='real_price')
plt.title("Apartment Prices by City")
plt.ylabel("Price (₽)")
plt.xticks(rotation=45)
plt.show()

# Average price per m²
df['price_per_m2'] = df['real_price'] / df['m2']
print(df.groupby('city')['price_per_m2'].mean().round(0))

---

license: mit