Chronify¶
This package implements validation, mapping, and storage of time series data in support of Python-based modeling packages.
Features¶
Stores time series data in any database supported by Ibis (DuckDB, SQLite, and Spark).
Supports data ingestion in a variety of file formats and configurations.
Supports efficient retrieval of time series through SQL queries.
Validates consistency of timestamps and resolution.
Provides mappings between different time configurations.
Supported Backends¶
Chronify uses Ibis for all database operations. The following backends are supported:
DuckDB (default)
SQLite (stores timestamps as text; casting or comparing them across column types is only millisecond-precise)
Apache Spark (via PySpark)
How to use this guide¶
Refer to How Tos for step-by-step instructions for creating store and ingesting data.
Refer to Tutorials examples of ingesting different types of data and mapping between time configurations.
Refer to Reference for API reference material.
Refer to Explanation for descriptions and behaviors of the time series store.