predictiveworks/cdap-spark

This project aims to implement the vision of Visual TS - Code-free orchestration of data pipelines (or workflow) to respond to analysis use cases for time series data. Working with time series data often suffers from missing entries. Interpolate is a CDAP computation plugin that addresses this issue for Apache Spark DataFrames. A frequent requirement for many time series analysis methods is that the data need to be stationary (i..e mean, variance and auto correlation structure do not change of time). For practical purposes, stationarity is usually determine from linear auto correlation functions (ACF).