User
zarrin-kzd opened this issue · 10 comments
Dear developer,
I want to use this tool.
I have seen the codes, and I have used the codes and the GUI.
In the first step, I tried to use the codes and GUI to get the results based on input data in the demo file.
Then I compared my results and the results in outputs of the demo file.
My fragility function results only covered the probability of exceedance of damage states for 0.01g to 1.44g for fragility_drift_PGA, and for the results of other IMs, this acceleration interval differed.
I need to know why this happens and how can I expand the X-Axis to get the results for 0.05g to 5g, according to the demo file??
Thanks in advance,
Zarrin
Dear Zarim,
Many thanks for the feedback, it is extremely valuable for us.
The demo files have been created based on the European Strong motion database. This database has records up to around 0.4g. The the tool has been designed to output values for imls between 0.5min(IML) and 2.0max(IML) (IML being computed from the ground motion records used in analyses).
I can also push an update to extend this range, but in the meantime you can edit the x_vec variable in the fragility_censored_cloud_analysis function to better suit your needs.
Sincerely
L.M.
Dear Luis,
My problem is in getting and plotting the resulted fragility curves.
I want to know that how I can get and plot the outputs in such a way that the X-axis limit is 5.0g for all IMs.
As you said, I have edited the x_vec variable in the fragility_censored_cloud_analysis function.
The tool has been edited to output values for imls between 0.5min(IML) and 4.0max(IML).
I have used the European Strong motion database and Capacity CSV files in the demo file.
Then I have run the software.
I have been attached some figures to this message so that you can better understand what I mean.
The top two figures are the results that I get from software for the "CR_LFINF-CDN_H4" building class.
The bottom two figures are your results for the same building class.
I need to know if there is any way to have the same X-Axis limit to get the results for 0.05g to 5g for all the curves (all building types), similar to what you did in your paper??
Thank you for your time and consideration.
Sincerely yours,
Zarrin
Good morning Zarim,
As you can see from your curves in SA(0.3) you are getting the same results. If you need to have the x_axis between 0.05g and 5g regardless of the IM you can edit the source code in the same line to x_vec=np.linspace(np.log(0.05),np.log(5),endpoint=True).
I do this sometimes for plotting reasons as well, simply to make the plots prettier. However, the reason why the source code is tied to the iml computed from the ground motion records and not to a fixed interval of IML is because 5g in PGA means a very different thing as 5g in SA(1.0s). A PGA of 5g in fact is about 1.5g in SA(1.0s).
Knowing this if you just need the plots you can apply the fix I told you above.
Cheers
L.M.
Dear Luis,
Thank you very much for the valuable information.
As you said, I want to have the x_axis between 0.05g and 5g for all building classes and all IMs to show the plots more regularly.
Your explanation was very helpful.
Best wishes,
Zarrin
Dear Luis,
I have a question about getting outputs from the software.
I need to know if there is a way to have the fragility parameters (median and logarithmic standard deviation) for all building types in the database using VMTK?
Best regards,
Zarrin
Good morning Zarrim,
thank you for your questions. Until now there was no way to compute the fragility parameters (because the cloud analysis doesn't need a regression). However i just pushed an update with a function get_logmean_logstddev that outputs the logmean and the logstddev of lognormal regression. The functions has 2 inputs, one an array of imls and another with the values of the probability of exceedance at the damage state of interest (or loss ratio in case you want to have the vulnerability curve).
You can add this function to your scripts to output the values you are interested in.
cheers
L.M>
Dear Luis,
Thank you for your response.
I have edited the python scripts as you have said.
Then I have run the software, but I didn't get the median and logarithmic standard deviation of lognormal distribution in outputs.
I need to get fragility parameters output with running Start_GUI.py and computation_scripts_example.py.
How can I edit these to get the median and logarithmic standard deviation of lognormal distribution in the outputs folder?
Best wishes,
Zarrin
Good afternoon Zarim,
I was under the impression you were computing using python's scripting capabilities. I have not altered the GUI. You can use the outputs of the GUI and the new function and write a simple script that reads the fragility csv and outputs the logmean and logstdev.
cheers
L.M.
Good afternoon Luis,
I'm an MSc student working with Matlab, and I am quite new to python.
I have written Matlab code to get the median and logarithmic standard deviation of the lognormal distribution.
I have computed these parameters for my fragility function results with Matlab code easily.
I sent two previous messages because you have said that It's good to send feedback.
Anyway, It's good that the non-professional users with python can get these parameters from The VMTK.
Best wishes,
Zarrin
Dear Zarrim,
I really appreciate your inputs. They will be super helpful to improve the tool's usability.
Please do keep sending feedback
Many thanks
L.M.