ts-flint - Time Series Library for PySpark

ts-flint is a collection of modules related to time series analysis for PySpark.

Reading Data with FlintContext

Reading and Writing Data shows how to read data into a ts.flint.TimeSeriesDataFrame, which provides additional time-series aware functionality.

>>> prices = (flintContext.read
...           .range(begin, end)
...           .uri(uri))

Manipulating and Analyzing Data

Manipulating and Analyzing Data describes the structure of ts.flint.TimeSeriesDataFrame, which is a time-series aware version of a pyspark.sql.DataFrame. Being time-series aware, it has optimized versions of some operations like joins, and also some new features like temporal joins. ts.flint.summarizers contains aggregation functions like EMAs.

>>> events.leftJoin(returns, tolerance='5d', key='id')


Indices and tables