Running conditional mean spectrum for different conditioning periods from a dstore of a completed classical PSHA

[Enrico Abcede]

Hello all,

I would like to calculate a conditional mean spectrum in an interactive dev. like Jupyter so I don't have to perform classical PSHA every time i want a CMS for a different conditioning period. However, I am unsure how to input the cross_correl and imt_ref since the conditional_spectrum.main only takes in the dstore as an argument, unlike case_mrd that takes multiple inputs.

Is there a way to update the oqParam of the dstore such that I can manually change the imt_ref and the cross_correl attributes?

I also welcome other ideas how to compute the CMS of other conditioning periods without having to run classical PSHA every time.

Cheers

[Michele Simionato]

Notice that the conditional spectrum calculator is meant to be run with the --hc option. The workflow is as follows:

1. run the first calculation and save its ID (for instance say it is 42):

2. run additional calculations starting from the first one (this will avoid having to recompute the classical part) and changing the parameters,

 for instance:

$ oq engine --run job.ini --hc 42 -p imt_ref="SA(0.1),cross_correlation=Bradley2012"

$ oq engine --run job.ini --hc 42 -p imt_ref="SA(0.2),cross_correlation=Bradley2012"

$ oq engine --run job.ini --hc 42 -p imt_ref="SA(0.1),cross_correlation=BakerJayaram2008"

$ oq engine --run job.ini --hc 42 -p imt_ref="SA(0.2),cross_correlation=BakerJayaram2008"

...

This is documented in the manual, see the section

https://docs.openquake.org/oq-engine/3.20/manual/user-guide/advanced/advanced-calculations.html#the-post-processing-framework-and-vector-valued-psha-calculations

and below.

            Michele Simionato