discontinuum 0.5

Estimate discontinuous timeseries from continuous covariates.

discontinuum

[!WARNING]
Experimental.

Overview

discontinuum is a middleware for developing Gaussian process (GP) models. Why might we want a middleware? GP's are a flexible and elegant approach to modeling dynamical systems for which we have sparse and uncertain observations. In this arena, simple GP models, specified in several lines of math, can often achieve state-of-the-art predictive performance. However, fitting GP's is numerically intense, $\mathcal{O}(n^3)$ complexity. They have several optimizations that take advantage of simplifying assumptions, different algorithms, or GPUs, but each has tradeoffs. Ideally, we could quickly write mathematical models, then run them on whichever "engine" is best suited for a particular problem.

Furthermore, most models include a lot of relatively standard utility functions for plotting, managing metadata, data pre-processing, and other "boiler plate." discontinum packages engines and utilities within a single ecosystem, such that creating a new model is just a matter of writing a little math without too much boilerplate.

Installation

pip install discontinuum

Models

loadset-gp

loadest-gp is Gaussian-process model for estimating river constituent time series, which borrows its namesake from the venerable LOAD ESTimator (LOADEST) software program. However, LOADEST has several serious limitations ---it's essentially a linear regression---and it has been all but replaced by the more flexible Weighted Regression on Time Discharge and Season (WRTDS), which allows the relation between target and covariate to vary through time. loadest-gp takes the WRTDS idea and reimplements it as a GP. Try it out in the loadest-gp demo.

rating-gp

rating-gp is a Gaussian-process model for estimate river flow from stage time series. Try it out in the rating-gp demo.

Engines

Currently, the only supported engines are the marginal likelihood implementation in pymc and gpytorch. Latent GP implementations could be added in the future. In general, the gpytorch implementation is faster and provides a lot a powerful features, like GPU support, whereas pymc is a more complete probabilistic-programming framework, which can be "friendlier" for certain use cases.

Roadmap

mindmap
  root((discontinuum))
    data providers
      USGS
      etc
    engines
      PyMC
      PyTorch
    utilities
      pre-processing
      post-processing
      plotting
    models
      loadest-gp
      rating-gp