Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.


Messages - fmk

Pages: [1] 2 3 ... 16
1
Regional Hazard Simulation (R2D, rWhale) / Re: R2D on Linux HPC
« on: October 27, 2023, 04:41:16 PM »
Amir,

It is possible to run on such a system if you can build both the backend and frontend. I am in the process of builidng a Dockerfile which you could use to see which packages are needed on your system and what the build process is. The Dockerfile uses Ubuntu18.04 as its base. The biggest issue I foresee is getting Qt version 5.15 installed. Failing to install the frontend would mean you could not use the GUI. The backend has a simpler build process and you could run the computational simulations there and possibly view results in python.


timeline for Dockerfile is probably next week (today being the 10/27/2023 for those others looking at this message). The Dockerfile will be in the R2DTool repo on github.

The purpose of the docker images is to run the frontend and backend on one of the TACC Frontera HPC compute nodes.

frank

2
Sorry there us no automated process. It was done manually.

3
I see you found a solution to install 2.0.0 instead of 2.0.5. thanks

4
Seth .. thanks .. I was having the same problems .. neither google or stackoverflow was of any  help .. I did not think of trying an earlier version.

5
Installation / Re: Conan in the CI
« on: August 23, 2023, 06:59:33 AM »
thanks. i have updated the appveyor files to use 1.60.1

6

Seth,

getting an image from all angles to better determine if there is a soft-story or not is problematic. StreetView will not provide all images necessary as for a majority of houses in urban areas one image from streetview is all that will be avail as neighbouring houses and not streets will be on the other 3 sides. An alternative we discussed briefly was to use drone imagery,  using drones that the NHERI Rapid facility provides for example, to capture all sides of the house. We just did not get far with that effort as the programmer looking after that effort left and nobody expressed much interest in that particular module.

7
Seth,

sorry about the problems. We moved from tensorflow to pytorch and not all the modules got revised. the Softstory is avail in the 2.0.5 release.
https://pypi.org/project/BRAILS/2.0.5/ . That release uses older python modules.

I am going through the process of fixing the documentation. In the meantime the modules and the doc for their usage with the current brails can best be found on the github dir for each module where the README.rst file displayed when using the browser shows how to use the module.
https://github.com/NHERI-SimCenter/BRAILS/tree/master/brails/modules

8
Uncertainty Quantification (quoFEM) / Re: Run QuoFEM with OpenSeesSP
« on: August 18, 2023, 08:29:14 PM »
if running locally .. you can (I believe) start OpenSeesSP w/o using the mpi command .. if this is the case replace the OpenSees exe that came with the app with OpenSeesSP and it will work fine.

9
Yes only 5, the number specified in the UQ tab,  would be used.

10
from the developer:

 
Kuanshi Zhong
Fri, May 19, 11:54 AM (5 days ago)
to me

Hi Frank,

We're using the UCERF3 - TD (time dependent).  The time dependence probability of a rupture in UCERF3-TD model is conditional on a time duration - so when output the mean annual rate, OpenSHA converts the time-dependent rate to time independent rate (please see more details in the "Example of Coverting Time-Dependent Probability into an Effective Time-Indpendent Rate" on the page 2 of this OpenSHA's tech note: https://opensha.org/resources/PSHA_Primer_v2_0.pdf).

I think you've directed the user to the "FetchOpenSHA.py" - he should be able to find that Line 218 invokes getMeanAnnualRate() in OpenSHA (https://github.com/opensha/opensha/blob/8a65cdc9ad27f5c3245688759f6c687ee95eff15/src/main/java/org/opensha/sha/earthquake/ProbEqkRupture.java#L57) to do the above conversion.  Just checked that the time span duration used is 30 years.

11
1. as for missing exe, did you download the mad or windows version?
2. i will ask the developer about UCERF to see if he knows as I am seeing nothing in the code that actually fetches the stuff from OpenSHA to answer the question., The code b.t.w is here:  https://github.com/NHERI-SimCenter/SimCenterBackendApplications/blob/master/modules/performRegionalEventSimulation/regionalGroundMotion/FetchOpenSHA.py

12
can you attach the files Kuanshi mentioned before: "if you'd like to share (1) one snapshot of the "Preferences" window (can be launched by clicking on "File"->"Preferences" in the top menu bar) and (2) the configuration file "EQHazardConfiguration.json" in the working directory ("C:\Users\emily\Documents\R2D\LocalWorkDir\HazardSimulation\Input") "
thanks

13
can you download the code in the preRelease folder. it should shown version #3.0.0. That image is for this new tool. People are testing the new tool for us and the doc has been updated for them.

14
We have just put out a new pre release for testing. In pre-release folder, has version #3.0.0,. In this new incarnation the earthquake selection has been moved to the tools menu-bar and as such you can use the tool w/o having to run the rest of the workflow.

can you send the ~/Documents/R2D/debug.log file to see if we can see what is up. if that fails possibly a zoom call will be needed.

15
Regional Hazard Simulation (R2D, rWhale) / Re: R2D Tool Error
« on: May 01, 2023, 02:09:53 AM »
look in your Documents/R2D/LocalWorkDir/tmp.SimCenter/Results .. there should be a log.txt file .. if there is a file, open it and look for the first ERROR .. if no file open the Documents/R2D/debug.log and see if you see an error .. if you cannot see anything, post the 2 files if they exist.

Pages: [1] 2 3 ... 16