Metadata-Version: 2.1
Name: azuremlftk
Version: 0.1.19032.1
Summary: "Microsoft Azure Machine Learning Forecasting Toolkit"
Home-page: https://aka.ms/aml-packages/forecasting
Author: "Microsoft Corporation"
Author-email: fcstcore@microsoft.com
License: UNKNOWN
Platform: UNKNOWN
Requires-Python: <3.7
Description-Content-Type: text/x-rst
Requires-Dist: dask (>=0.17.2)
Requires-Dist: dill
Requires-Dist: distributed (==1.23.1)
Requires-Dist: h5py
Requires-Dist: html5lib
Requires-Dist: joblib
Requires-Dist: jsonschema (>=2.6.0)
Requires-Dist: keras
Requires-Dist: matplotlib (>=3.0.2)
Requires-Dist: numexpr
Requires-Dist: numpy (>=1.11.3)
Requires-Dist: pandas (<=0.23.4,>=0.21.0)
Requires-Dist: patsy
Requires-Dist: requests (>=2.18.4)
Requires-Dist: scikit-learn (>=0.19.1)
Requires-Dist: scipy
Requires-Dist: six (>=1.11.0)
Requires-Dist: statsmodels (>=0.9.0)
Requires-Dist: tensorflow (>=1.10.0)
Requires-Dist: azureml-core
Requires-Dist: azureml-telemetry
Requires-Dist: azure-storage-blob (>=1.1.0)
Requires-Dist: azureml-train-core
Provides-Extra: databricks
Requires-Dist: pyOpenSSL (==16.0.0) ; extra == 'databricks'
Requires-Dist: cryptography (==1.5) ; extra == 'databricks'
Requires-Dist: ipython (==2.2.0) ; extra == 'databricks'
Requires-Dist: numpy (==1.11.3) ; extra == 'databricks'
Provides-Extra: sql_support
Requires-Dist: sqlparse ; extra == 'sql_support'
Requires-Dist: pyodbc ; extra == 'sql_support'
Requires-Dist: sqlalchemy ; extra == 'sql_support'

Azuremlftk is the Azure ML Package for Forecasting, a library of models, data transforms and evaluators common in forecasting scenarios.

Some of the things it provides:
* data imputation transforms
* rolling window and lagging/leading transforms
* timestamp featurization
* validators for doing cross-validation correctly with time series
* standard "univariate" models" like ARIMA, ETS and Seasonal Naive
* forecasting models that seamlessly convert forecasting to regression and accept external regressors
* grouping multiple time series to learn a single model

All models and transforms are presented with uniform scikit-like interfaces.

Reference documentation https://azuremlpackages.blob.core.windows.net/forecasting/release1902/documentation/index.html

Sample notebooks https://github.com/Azure-Samples/MachineLearningSamples-Notebooks/tree/master/domain-packages/forecasting




