electricsheepafrica/africa-synth-retail-and-ecommerce-refund-and-return-logs-nigeria

⚠️ Synthetic dataset — Parameterized from published SSA literature, not real observations. Not suitable for empirical analysis or policy inference.

Refund And Return Logs

Dataset Description

Comprehensive refund and return logs for Nigerian retail and e-commerce analysis

Dataset Information

• Category: Sales and Transactions
• Industry: Retail & E-Commerce
• Country: Nigeria
• Format: CSV, Parquet
• Rows: 150,000
• Columns: 11
• Date Generated: 2025-10-06
• Location: data/refund_and_return_logs/
• License: GPL

Schema

Column	Type	Sample Values
return_id	String	RET0000000
order_id	String	ORD7712767
customer_id	String	CUST726777
return_date	String	2024-01-19
product_id	String	PRD87085
return_reason	String	late_delivery
refund_amount_ngn	Float	124407.43
refund_method	String	bank_transfer
processing_time_days	Integer	7
return_status	String	approved
restocking_fee_ngn	Float	0.0

Sample Data

 return_id   order_id customer_id return_date product_id     return_reason  refund_amount_ngn    refund_method  processing_time_days return_status  restocking_fee_ngn
RET0000000 ORD7712767  CUST726777  2024-01-19   PRD87085     late_delivery          124407.43    bank_transfer                     7      approved                 0.0
RET0000001 ORD6474212  CUST447076  2024-03-26   PRD84276      changed_mind          195071.65    bank_transfer                    10      approved                 0.0
RET0000002 ORD4097385  CUST865931  2024-09-12   PRD14165 defective_product          169538.24 original_payment                     8      approved                 0.0

Use Cases

• Data analysis and insights
• Machine learning model training
• Business intelligence
• Research and education
• Predictive analytics

Nigerian Context

This dataset incorporates authentic Nigerian retail and e-commerce characteristics:

E-Commerce Platforms

• Jumia (35% market share) - Leading marketplace
• Konga (25% market share) - Major competitor
• Jiji (20% market share) - Classifieds platform
• PayPorte, Slot, and other platforms

Physical Retail

• Shoprite, Spar, Game - Major supermarket chains
• Slot, Pointek - Electronics retailers
• Mr Price - Fashion retail
• Traditional markets: Balogun Market, Computer Village

Payment Methods

• Cash on Delivery (45%) - Most popular
• Bank Transfer (25%)
• Debit Card (15%)
• USSD (8%)
• Mobile Money (5%)
• Credit Card (2%)

Logistics & Delivery

• GIG Logistics - Nationwide coverage
• Kwik Delivery - Fast urban delivery
• DHL, FedEx - International and express
• Red Star Express - Nationwide courier
• Local dispatch riders

Geographic Coverage

Major Nigerian cities including:

• Lagos - Commercial capital, highest retail density
• Abuja - Federal capital, high e-commerce penetration
• Kano - Northern commercial hub
• Port Harcourt - Oil city, strong purchasing power
• Ibadan - Large urban market
• Plus 10+ other major cities

Products & Categories

• Electronics: Tecno, Infinix, Samsung phones; laptops, TVs
• Fashion: Ankara fabric, Agbada, Kaftan, sneakers
• Groceries: Rice (50kg bags), Garri, Palm Oil, Indomie
• Beauty: Shea butter, Black soap, hair extensions
• Home: Generators, inverters, solar panels

Currency & Pricing

• Currency: Nigerian Naira (NGN, ₦)
• Exchange Rate: ~₦1,500/USD
• Price Ranges: Realistic Nigerian market prices
• Time Zone: West Africa Time (WAT, UTC+1)

File Formats

CSV

data/refund_and_return_logs/nigerian_retail_and_ecommerce_refund_and_return_logs.csv

Parquet (Recommended)

data/refund_and_return_logs/nigerian_retail_and_ecommerce_refund_and_return_logs.parquet

Nigerian Retail and E-Commerce - Loading the Dataset

Hugging Face Datasets

from datasets import load_dataset

# Load dataset
dataset = load_dataset("electricsheepafrica/nigerian_retail_and_ecommerce_refund_and_return_logs")

# Convert to pandas
df = dataset['train'].to_pandas()

print(f"Loaded {len(df):,} rows")

Pandas (Direct)

import pandas as pd

# Load CSV
df = pd.read_csv('data/refund_and_return_logs/nigerian_retail_and_ecommerce_refund_and_return_logs.csv')

# Load Parquet (recommended for large datasets)
df = pd.read_parquet('data/refund_and_return_logs/nigerian_retail_and_ecommerce_refund_and_return_logs.parquet')

PyArrow

import pyarrow.parquet as pq

# Load Parquet
table = pq.read_table('data/refund_and_return_logs/nigerian_retail_and_ecommerce_refund_and_return_logs.parquet')
df = table.to_pandas()

Data Quality

• ✅ Realistic Distributions: Based on Nigerian retail patterns
• ✅ No Missing Critical Fields: Complete core data
• ✅ Proper Data Types: Appropriate types for each column
• ✅ Consistent Naming: Clear, descriptive column names
• ✅ Nigerian Context: Authentic local characteristics
• ✅ Production Scale: Suitable for real-world applications

Ethical Considerations

• This is synthetic data generated for research and development
• No real customer data or personally identifiable information
• Designed to reflect realistic patterns without privacy concerns
• Safe for public use, testing, and education

License

GPL License - General Public License

This dataset is free to use for:

• Research and academic purposes
• Commercial applications
• Educational projects
• Open source development

Citation

@dataset{nigerian_retail_refund_and_return_logs_2025,
  title={Refund And Return Logs},
  author={Electric Sheep Africa},
  year={2025},
  publisher={Hugging Face},
  howpublished={\url{https://huggingface.co/datasets/electricsheepafrica/nigerian-retail-refund-and-return-logs}}
}

Related Datasets

This dataset is part of the Nigerian Retail & E-Commerce Datasets collection, which includes 42 datasets covering:

• Customer & Shopper Data
• Sales & Transactions
• Product & Inventory
• Marketing & Engagement
• Operations & Workforce
• Pricing & Revenue
• Customer Support
• Emerging & Advanced Technologies

Browse all datasets: https://huggingface.co/electricsheepafrica

Updates & Maintenance

• Version: 1.0
• Last Updated: 2025-10-06
• Maintenance: Active
• Issues: Report via Hugging Face discussions

Contact

For questions, feedback, or collaboration:

• Hugging Face: electricsheepafrica
• Issues: Open a discussion on the dataset page
• General Inquiries: Via Hugging Face profile

---

Part of the Nigerian Industry Datasets Initiative
Building comprehensive, authentic datasets for African markets.