shacharu/panik-full

Stock Market OHLCV Dataset

Dataset Description

This dataset contains historical OHLCV (Open, High, Low, Close, Volume) data for US stock market symbols, processed and structured for machine learning and financial research.

Dataset Summary

• Total Rows: 27,869,709
• Number of Symbols: 8818
• Date Range: 1962-01-02 to 2025-08-29
• Sample Symbols: MEMS, TEXN, ATXG, FDBC, PAWZ, LII, INLF, JUNS, MEDX, QVMM
• Created: 2025-09-02

Schema

The dataset includes the following columns:

• symbol: Stock ticker symbol (string)
• date: Trading date (date32)
• open, high, low, close: Price data (float32)
• volume: Trading volume (int64)
• exchange: Stock exchange classification
• liquidity_tier: Liquidity classification (Institutional, Liquid, Opportunistic, Watchlist)
• market_cap: Market capitalization category
• sector: Industry sector
• data_source: Source of the data (yahoo, stooq, etc.)
• last_updated: Last update timestamp

Usage

from datasets import load_dataset

# Load the dataset
dataset = load_dataset("shacharu/panik-full")

# Access the data
print(f"Dataset size: {len(dataset['train']):,} rows")
print(f"Columns: {list(dataset['train'].features.keys())}")

# Filter by symbol
aapl_data = dataset['train'].filter(lambda x: x['symbol'] == 'AAPL')

# Convert to pandas for analysis
import pandas as pd
df = dataset['train'].to_pandas()

Data Sources

Primary data sources include:

• Yahoo Finance
• Stooq
• Other financial data providers

License

MIT License - Free for research and commercial use.

Citation

@dataset{panik_stock_ohlcv_2024,
  author = {Panik Trading Research},
  title = {Stock Market OHLCV Dataset},
  year = {2024},
  url = {https://huggingface.co/datasets/shacharu/panik-full}
}

Contact

For questions or issues, please open an issue in the dataset repository.