search

antoinecarme / pyaf

PyAF is an Open Source Python library for Automatic Time Series Forecasting built on top of popular pydata modules.
BSD 3-Clause "New" or "Revised" License
459 stars 72 forks source link

Add ability to specify confidence interval #60

Closed mgiangreco closed 7 years ago

mgiangreco commented 7 years ago

Great package. Would be nice to add the ability to specify the confidence interval used for generating the Upper and Lower bound predictions (e.g. 90%, 75%, 50%, etc.)

antoinecarme commented 7 years ago

@mgiangreco

Thanks.

Sure , it is possible to add the possibility to customize the confidence interval. 95% is used by default, and for the majority of statisticians it covers the most interesting use cases.

Will not have the necessary bandwidth for the next two months ... but I add it to my planning. If you are interested in adding it (expect more test scripts than code ;), you are welcome.

antoinecarme commented 7 years ago

Until a fix is implemented, a work around is to change the value of 1.96 by the desired z-value (file TS/TimeSeriesModel.py).

see https://en.wikipedia.org/wiki/Confidence_interval for the z-values for 90%, 75%, 50% etc.