Skip to main content

A package with helper scripts for complex DataRobot AutoTS use cases

Project description

datarobot_ts_helpers package

A library of helper scripts to support complex time-series modeling using DataRobot AutoTS software

Authors

Justin Swansburg, Jarred Bultema, Jess Lin

Description

The modeling of large scale time-series problems is possible directly within DataRobot software via the GUI or via R or Python modeling APIs. While the software is capable of modeling up to 1 million series per project and applying state of the art modeling techniques, often there is motivation to model aspects of a data science problem across multiple DataRobot projects. Motivation for this may include a desire to externally cluster similar series, apply different data manipulations or corrections, utilize different data sources, apply different differencing strategies, utilize different Feature Derivation Windows, or investigate different Forecast Distance ranges. Regardless of the reasons, internally we have found that performance can often be improved on large or complex time-series use cases by breaking a large, challenging problem into smaller pieces and modeling each of those pieces separately.

This is feasible directly using the R or Python modeling APIs, but the challenge quickly becomes one of software engineering and logistics to manage, compare, and store outputs of numerous projects that are part of a single use-case. The purpose of the ts_helpers package is to automate this logistical challenge and allow the DataRobot user to focus on applying different approaches to solve their use case, rather than focusing on the less interesting aspects of the problem.

Contents

This python package contains numerous functions to enable the user to easily scale from one to thousands of DataRobot projects starting with data preparation and continuing through modeling, model evaluation, iterative performance improvements, visualization of results, deployment of models, and serving ongoing predictions.

A detailed Table of Contents describes all functions present and the documentation string for each function. Detailed tutorials are also available to demonstrate the use of this ts_helpers package and all of the functions contained within.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distributions

No source distribution files available for this release.See tutorial on generating distribution archives.

Built Distribution

datarobot_ts_helpers-0.1.5-py3-none-any.whl (49.9 kB view details)

Uploaded Python 3

File details

Details for the file datarobot_ts_helpers-0.1.5-py3-none-any.whl.

File metadata

  • Download URL: datarobot_ts_helpers-0.1.5-py3-none-any.whl
  • Upload date:
  • Size: 49.9 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/3.3.0 pkginfo/1.5.0.1 requests/2.22.0 setuptools/51.1.2 requests-toolbelt/0.9.1 tqdm/4.48.0 CPython/3.7.4

File hashes

Hashes for datarobot_ts_helpers-0.1.5-py3-none-any.whl
Algorithm Hash digest
SHA256 3552b7b84699c198d5ca4c808c7955154a2a4605ee4a5e566ec20fa0c4c51e8b
MD5 2f06385153c3e50833abd40a7265d5af
BLAKE2b-256 aa6ba244f9b888d6734ce5ef95abba43ce0c1659a28edbaa1e63e068a5d109c7

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page