Closed emongold closed 3 years ago
Thanks Emily for checking this. It’s interesting. The verification tables we have were obtained from the Matlab code provided as an electronic supplement to the EQS paper (see https://journals.sagepub.com/doi/suppl/10.1193/070913EQS198M). Of course there is a possibility that we (actually I) did a mistake. I wonder if you also used this code for comparison.
Comparing the results of the test file for abrahamson_2014 to the GMPE master spreadsheet and to this equivalent Matlab code, the OpenQuake version is inconsistent with the results from the other two.
-- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/gem/oq-engine/issues/7099
Marco, I had not seen this verification code, but when I run it the results seem to align with our (Baker Research Group) Matlab function and not the OpenQuake results. Emily
From: Marco Pagani @.> Sent: Thursday, August 19, 2021 12:33 PM To: gem/oq-engine @.> Cc: Emily Louise Mongold @.>; Author @.> Subject: Re: [gem/oq-engine] Abrahamson 2014 model bug (#7099)
Thanks Emily for checking this. It’s interesting. The verification tables we have were obtained from the Matlab code provided as an electronic supplement to the EQS paper (see https://journals.sagepub.com/doi/suppl/10.1193/070913EQS198M). Of course there is a possibility that we (actually I) did a mistake. I wonder if you also used this code for comparison.
Comparing the results of the test file for abrahamson_2014 to the GMPE master spreadsheet and to this equivalent Matlab code, the OpenQuake version is inconsistent with the results from the other two.
-- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/gem/oq-engine/issues/7099
MARCO PAGANI | Seismic Hazard Team Lead | Skype mm.pagani | +39-0382-5169863 GLOBAL EARTHQUAKE MODEL | working together to assess risk GEM - globalquakemodel.org | T - @GEMwrld | F - GEMwrld
“I do not know anything about luck, apart from that the more I practice, the luckier I get” Ingemar Stenmark
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/gem/oq-engine/issues/7099#issuecomment-902184618, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AR54OAIWA3TBUS33UVASCC3T5VL6ZANCNFSM5COSS3DA. Triage notifications on the go with GitHub Mobile for iOShttps://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Androidhttps://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email.
Thanks for checking this. If so, we will need to look into it. Any hints on the conditions (i.e. a specific scenario or combination of explanatory variables) for which you observed these differences?
Marco, I had not seen this verification code, but when I run it the results seem to align with our (Baker Research Group) Matlab function and not the OpenQuake results. Emily
From: Marco Pagani @.> Sent: Thursday, August 19, 2021 12:33 PM To: gem/oq-engine @.> Cc: Emily Louise Mongold @.>; Author @.> Subject: Re: [gem/oq-engine] Abrahamson 2014 model bug (#7099)
Thanks Emily for checking this. It’s interesting. The verification tables we have were obtained from the Matlab code provided as an electronic supplement to the EQS paper (see https://journals.sagepub.com/doi/suppl/10.1193/070913EQS198M). Of course there is a possibility that we (actually I) did a mistake. I wonder if you also used this code for comparison.
Comparing the results of the test file for abrahamson_2014 to the GMPE master spreadsheet and to this equivalent Matlab code, the OpenQuake version is inconsistent with the results from the other two.
-- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/gem/oq-engine/issues/7099
MARCO PAGANI | Seismic Hazard Team Lead | Skype mm.pagani | +39-0382-5169863 GLOBAL EARTHQUAKE MODEL | working together to assess risk GEM - globalquakemodel.org | T - @GEMwrld | F - GEMwrld
“I do not know anything about luck, apart from that the more I practice, the luckier I get” Ingemar Stenmark
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/gem/oq-engine/issues/7099#issuecomment-902184618, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AR54OAIWA3TBUS33UVASCC3T5VL6ZANCNFSM5COSS3DA. Triage notifications on the go with GitHub Mobile for iOShttps://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Androidhttps://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email.
-- You are receiving this because you commented. Reply to this email directly or view it on GitHub: https://github.com/gem/oq-engine/issues/7099#issuecomment-903009966
Marco, I found that errors were larger with larger magnitude, smaller Rrup distances, lower Vs30 values and large Z1.0. I was not able to find differences in the scripts when I went through. I just compared the mean values for California from the test file. Attached is my Matlab test script (compare_ask.m) along with all the functions you should need for it to run, in case that helps. Emily
From: Marco Pagani @.> Sent: Saturday, August 21, 2021 12:35 AM To: gem/oq-engine @.> Cc: Emily Louise Mongold @.>; Author @.> Subject: Re: [gem/oq-engine] Abrahamson 2014 model bug (#7099)
Thanks for checking this. If so, we will need to look into it. Any hints on the conditions (i.e. a specific scenario or combination of explanatory variables) for which you observed these differences?
Marco, I had not seen this verification code, but when I run it the results seem to align with our (Baker Research Group) Matlab function and not the OpenQuake results. Emily
From: Marco Pagani @.> Sent: Thursday, August 19, 2021 12:33 PM To: gem/oq-engine @.> Cc: Emily Louise Mongold @.>; Author @.> Subject: Re: [gem/oq-engine] Abrahamson 2014 model bug (#7099)
Thanks Emily for checking this. It’s interesting. The verification tables we have were obtained from the Matlab code provided as an electronic supplement to the EQS paper (see https://journals.sagepub.com/doi/suppl/10.1193/070913EQS198M). Of course there is a possibility that we (actually I) did a mistake. I wonder if you also used this code for comparison.
Comparing the results of the test file for abrahamson_2014 to the GMPE master spreadsheet and to this equivalent Matlab code, the OpenQuake version is inconsistent with the results from the other two.
-- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/gem/oq-engine/issues/7099
MARCO PAGANI | Seismic Hazard Team Lead | Skype mm.pagani | +39-0382-5169863 GLOBAL EARTHQUAKE MODEL | working together to assess risk GEM - globalquakemodel.org | T - @GEMwrld | F - GEMwrld
“I do not know anything about luck, apart from that the more I practice, the luckier I get” Ingemar Stenmark
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/gem/oq-engine/issues/7099#issuecomment-902184618, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AR54OAIWA3TBUS33UVASCC3T5VL6ZANCNFSM5COSS3DA. Triage notifications on the go with GitHub Mobile for iOShttps://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Androidhttps://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email.
-- You are receiving this because you commented. Reply to this email directly or view it on GitHub: https://github.com/gem/oq-engine/issues/7099#issuecomment-903009966
MARCO PAGANI | Seismic Hazard Team Lead | Skype mm.pagani | +39-0382-5169863 GLOBAL EARTHQUAKE MODEL | working together to assess risk GEM - globalquakemodel.org | T - @GEMwrld | F - GEMwrld
“I do not know anything about luck, apart from that the more I practice, the luckier I get” Ingemar Stenmark
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/gem/oq-engine/issues/7099#issuecomment-903075858, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AR54OAM2QEEMM45BOTNMLULT55JNFANCNFSM5COSS3DA. Triage notifications on the go with GitHub Mobile for iOShttps://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Androidhttps://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email.
Emily, the attachment did now make it through GitHub. Pls send it at @.*** Thanks again, Marco
Marco, I found that errors were larger with larger magnitude, smaller Rrup distances, lower Vs30 values and large Z1.0. I was not able to find differences in the scripts when I went through. I just compared the mean values for California from the test file. Attached is my Matlab test script (compare_ask.m) along with all the functions you should need for it to run, in case that helps. Emily
From: Marco Pagani @.> Sent: Saturday, August 21, 2021 12:35 AM To: gem/oq-engine @.> Cc: Emily Louise Mongold @.>; Author @.> Subject: Re: [gem/oq-engine] Abrahamson 2014 model bug (#7099)
Thanks for checking this. If so, we will need to look into it. Any hints on the conditions (i.e. a specific scenario or combination of explanatory variables) for which you observed these differences?
Marco, I had not seen this verification code, but when I run it the results seem to align with our (Baker Research Group) Matlab function and not the OpenQuake results. Emily
From: Marco Pagani @.> Sent: Thursday, August 19, 2021 12:33 PM To: gem/oq-engine @.> Cc: Emily Louise Mongold @.>; Author @.> Subject: Re: [gem/oq-engine] Abrahamson 2014 model bug (#7099)
Thanks Emily for checking this. It’s interesting. The verification tables we have were obtained from the Matlab code provided as an electronic supplement to the EQS paper (see https://journals.sagepub.com/doi/suppl/10.1193/070913EQS198M). Of course there is a possibility that we (actually I) did a mistake. I wonder if you also used this code for comparison.
Comparing the results of the test file for abrahamson_2014 to the GMPE master spreadsheet and to this equivalent Matlab code, the OpenQuake version is inconsistent with the results from the other two.
-- You are receiving this because you are subscribed to this thread. Reply to this email directly or view it on GitHub: https://github.com/gem/oq-engine/issues/7099
MARCO PAGANI | Seismic Hazard Team Lead | Skype mm.pagani | +39-0382-5169863 GLOBAL EARTHQUAKE MODEL | working together to assess risk GEM - globalquakemodel.org | T - @GEMwrld | F - GEMwrld
“I do not know anything about luck, apart from that the more I practice, the luckier I get” Ingemar Stenmark
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/gem/oq-engine/issues/7099#issuecomment-902184618, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AR54OAIWA3TBUS33UVASCC3T5VL6ZANCNFSM5COSS3DA. Triage notifications on the go with GitHub Mobile for iOShttps://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Androidhttps://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email.
-- You are receiving this because you commented. Reply to this email directly or view it on GitHub: https://github.com/gem/oq-engine/issues/7099#issuecomment-903009966
MARCO PAGANI | Seismic Hazard Team Lead | Skype mm.pagani | +39-0382-5169863 GLOBAL EARTHQUAKE MODEL | working together to assess risk GEM - globalquakemodel.org | T - @GEMwrld | F - GEMwrld
“I do not know anything about luck, apart from that the more I practice, the luckier I get” Ingemar Stenmark
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/gem/oq-engine/issues/7099#issuecomment-903075858, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AR54OAM2QEEMM45BOTNMLULT55JNFANCNFSM5COSS3DA. Triage notifications on the go with GitHub Mobile for iOShttps://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Androidhttps://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email.
-- You are receiving this because you commented. Reply to this email directly or view it on GitHub: https://github.com/gem/oq-engine/issues/7099#issuecomment-903308459
MARCO PAGANI | Seismic Hazard Team Lead | Skype mm.pagani | +39-0382-5169863 GLOBAL EARTHQUAKE MODEL | working together to assess risk GEM - globalquakemodel.org | T - @GEMwrld | F - GEMwrld
“I do not know anything about luck, apart from that the more I practice, the luckier I get” Ingemar Stenmark
Annoying ... marco.pagani at globalquakemodel.org
Marco, I tried to send it and my message was blocked. Here is a OneDrive link that should work: [https://outlook-2.cdn.office.net/assets/mail/file-icon/png/zip_16x16.png] OQ Comparison.ziphttps://office365stanford-my.sharepoint.com/:u:/g/personal/emongold_stanford_edu/ETWvvmRUq6pLnRmCBW8gi8ABDnTg8zotzo_Fyb2Uf5o_xw?email=marco.pagani%40globalearthquakemodel.org&e=CLxjR1. Emily
From: Marco Pagani @.> Sent: Tuesday, August 24, 2021 5:45 AM To: gem/oq-engine @.> Cc: Emily Louise Mongold @.>; Author @.> Subject: Re: [gem/oq-engine] Abrahamson 2014 model bug (#7099)
Annoying ... marco.pagani at globalquakemodel.org
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHubhttps://github.com/gem/oq-engine/issues/7099#issuecomment-904607822, or unsubscribehttps://github.com/notifications/unsubscribe-auth/AR54OAO6ZC5KH4W45Q6Y6BDT6OH5JANCNFSM5COSS3DA. Triage notifications on the go with GitHub Mobile for iOShttps://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Androidhttps://play.google.com/store/apps/details?id=com.github.android&utm_campaign=notification-email.
Hi Emily, I did not manage to get the files you shared since the email address you used to share the files was misspelt. I did an update of our verification tables and checked the model (https://github.com/gem/oq-engine/pull/7136). Overall the match seems acceptable. Pls, let me know if this work also on your side. Thanks
Comparing the results of the test file for abrahamson_2014 to the GMPE master spreadsheet and to this equivalent Matlab code, the OpenQuake version is inconsistent with the results from the other two.