Deaggregation calculation duration

[Marty Zaleski]

Hello all. Wondering about how long it should take to complete a deaggregation calculation. I'm in the midst of deaggregating two parameters (PGA and SA02), each at three hazard levels (1:475, 1:2475, 1:10,000). I'm deaggregating magnitude, distance, and magnitude-distance. I'm running the calculation on my laptop. It's been churning away for about 18 hours now. Is this the kind of thing that should take a day? a week? a month?

[Michele Simionato]

On Friday, May 31, 2019 at 12:34:01 AM UTC+2, Marty Zaleski wrote:

Hello all. Wondering about how long it should take to complete a deaggregation calculation. I'm in the midst of deaggregating two parameters (PGA and SA02), each at three hazard levels (1:475, 1:2475, 1:10,000). I'm deaggregating magnitude, distance, and magnitude-distance. I'm running the calculation on my laptop. It's been churning away for about 18 hours now. Is this the kind of thing that should take a day? a week? a month?

Perhaps a month, given that you are using the Canada model. We are very well aware of the performance issues of the disaggregation calculator. What you can do for the moment is to reduce the calculation to a single realization.

We would be very happy to start a project to improve the performance of the disaggregation calculator, we are just waiting for people wanting to sponsor it. If interested, please contact us.

            Michele

[Marty Zaleski]

Thank you for the reply. I'm not sure how to bring it down to one realization. What I'm really looking for is a magnitude-distance deaggregation of PGA and Sa(0.2) at the 1:10,000 exceedance probability level for one site. Could you please share a job.ini file that will make that work? Or at least share the calculation and deaggregation lines?

[Michele Simionato]

On Friday, May 31, 2019 at 6:15:31 AM UTC+2, Marty Zaleski wrote:

Thank you for the reply. I'm not sure how to bring it down to one realization. What I'm really looking for is a magnitude-distance deaggregation of PGA and Sa(0.2) at the 1:10,000 exceedance probability level for one site. Could you please share a job.ini file that will make that work? Or at least share the calculation and deaggregation lines?

Currently the only way to reduce the number of realizations is to manually edit the gsim logic tree file and the source model logic tree file, which is not particularly convenient. Also, you must decide in advance which is the realization you want and the engine is not helping, even if it could.

         MIchele 

[Atif Rasheed Baloch]

Hi,

      Please help or teach me to calculate the mean and model from the results obtained from

openQuake. I'm new in Deaggration analysis .

Thanks

Atif Rasheed

[Sylvanus Ahulu]

Marty Zaleski & Atif Rasheed,

     Please have you got a solution to the calculation of mean and model for deaaggregation? I need help !!

Sylvanus

On Thursday, 30 May 2019 23:34:01 UTC+1, Marty Zaleski wrote:

Hello all. Wondering about how long it should take to complete a deaggregation calculation. I'm in the midst of deaggregating two parameters (PGA and SA02), each at three hazard levels (1:475, 1:2475, 1:10,000). I'm deaggregating magnitude, distance, and magnitude-distance. I'm running the calculation on my laptop. It's been churning away for about 18 hours now. Is this the kind of thing that should take a day? a week? a month?