metadata

license: apache-2.0
task_categories:
  - time-series-forecasting
tags:
  - finance
  - crypto
  - economics
  - trading
  - blockchain
  - quantitative-analysis
  - machine-learning
  - deep-learning
  - time-series
  - sequence-modeling
  - price-prediction
  - market-analysis
  - investment-strategies
  - technical-indicators
  - historical-data-analysis
language:
  - en
multilinguality:
  - monolingual
pretty_name: CryptoData Dataset

CryptoData Dataset

The CryptoData dataset is a comprehensive collection of cryptocurrency market data, designed to support various analyses, including price prediction, market trend analysis, and the study of the impact of various indicators on cryptocurrency prices.

This dataset has been configured to provide flexibility in selecting specific types of market data through the use of different dataset configurations. Depending on the analysis needs, users can select one of the available configurations to load data tailored to their requirements.

Available Configurations:

Default: Includes open, high, low, close, and volume for each cryptocurrency market and date.
Close: Focuses on the close price and volume of each cryptocurrency market and date, optimized for simplicity and analyses centered on closing prices.
Indicators: Expands upon the default configuration by including technical indicators such as RSI (Relative Strength Index), SMA (Simple Moving Average), and EMA (Exponential Moving Average), aimed at more advanced technical analyses.
Sequences: Specifically designed for sequence prediction tasks, this configuration provides sequences of market data alongside the corresponding prediction targets, facilitating the development of models for future price prediction.

How to Use:

Below are Python code snippets demonstrating how to load the CryptoData dataset with each configuration. Before running the snippets, ensure you have the datasets library from Hugging Face installed.

from datasets import load_dataset

# Load the default configuration
dataset_default = load_dataset("crypto_data", config_name="default")

# Load the 'close' configuration
dataset_close = load_dataset("crypto_data", config_name="close")

# Load the 'indicators' configuration
dataset_indicators = load_dataset("crypto_data", config_name="indicators")

# Load the 'sequences' configuration
dataset_sequences = load_dataset("crypto_data", config_name="sequences")

Dataset Structure:

market: The cryptocurrency market (e.g., "BTC-USD").
date/time: The date or time of the data point.
open, high, low, close: Open, high, low, and close prices for the cryptocurrency.
volume: The volume of transactions.
rsi, sma, ema: Technical indicators including Relative Strength Index, Simple Moving Average, and Exponential Moving Average (available in the indicators configuration).
sequence, prediction: Arrays of historical data and the corresponding future data to predict (available in the sequences configuration).

Important Notes:

This dataset is for academic and research purposes only. Ensure compliance with any usage restrictions set by the data provider.
When using technical indicators in your analysis, be aware that these indicators alone may not provide a complete picture of market dynamics.
The sequences configuration requires significant preprocessing, including the calculation of technical indicators and the formation of sequences. This configuration is best suited for those with experience in time series analysis and deep learning.

Citation and Acknowledgments:

This dataset is made available for public use by the cryptocurrency research community. While there is no specific citation for this dataset, users are encouraged to reference the dataset's URL and the corresponding author's contributions.

Homepage: CryptoData Dataset on Hugging Face

For any questions or issues with the dataset, please raise an issue on the repository hosting the dataset.