Hello all,
Associated with the latest merged pull request in RAVEN for an AutoARMA algorithm (#2309) are some changes that will affect users:
- The input for the ARMA TSA algorithm has been switched from “SignalLag” and “NoiseLag” back to “P” and “Q” to conform with how the terms are referenced in literature
- Old RAVEN XML scripts used to train a SyntheticHistory ROM that use the ARMA algorithm will need to be updated with the new input names; see
tests/framework/ROM/TimeSeries/SyntheticHistory/arma.xml for an example of the new syntax
- Conversion scripts have been provided in
scripts/conversionScripts/SyntheticHistory_ARMA_lag.py; for more information on running the conversion scripts, see the README at
https://github.com/idaholab/raven/tree/devel/scripts/conversionScripts
- A new python package (statsforecast) has been added to the dependencies list which will require re-installation of the
raven_libraries python environment.
- ./scripts/establish_conda_env.sh --install will have to be run the next time a user updates RAVEN from git.
- See the
wiki for more information on updating RAVEN
Best,
Gabriel
Gabriel J. Soto Gonzalez, PhD
(he/him/his)
Modeling and Simulation Specialist | Digital Reactor Technology & Development
(C160)
Gabriel.So...@inl.gov |
786-712-8664
Idaho National Laboratory
_______________________________
