---
license: cc-by-nc-sa-4.0
tags:
- finance
- time-series
---
This dataset consists of snapshots of limit order books and trades for BTC/USD (i.e. the Bitcoin / US dollars currency pair) from May 31, 2018 9:55 pm (UTC) through September 30, 2018 9:59 pm (UTC) from the Bitstamp exchange (https://www.bitstamp.net).

The data has been collected by Aisot Technologies AG, Zürich (www.aisot.com). Trade data is on a millisecond frequency. Limit order book snapshots are on minute frequency, with aggregated amounts for each price level with depth up to 5000 for each bid/ask side. For more information about the dataset, please refer to the citation below. 


The data is provided “as is” without any warranties. A short approval process is required before accessing the data. 
By accessing the dataset, you accept to not disseminate it elsewhere and to adhere to the 
cc-by-nc-sa-4.0 license agreement. 
Note, we approve requests with full name (first and last name) and email only.

How to cite the dataset: Antulov-Fantulin, N., Guo, T. & Lillo, F. (2021). “Temporal mixture ensemble models for probabilistic forecasting of intraday cryptocurrency volume.” In: Decisions Econ. Finan. 44, pp. 905–940. https://doi.org/10.1007/s10203-021-00344-9