anomalous Standard Deviation

This forum is dedicated to discuss all problems and suggestions related to the inversion software
Post Reply
luigiV
Posts: 121
Joined: Fri Jul 17, 2020 9:28 am

anomalous Standard Deviation

Post by luigiV »

Hi Marc,
manually applying a standard deviation of 10% to a dispersion curve I get a bar that seems absolutely incorrect to me. What could be the problem? I recently installed windows 11....is that going to be it? Is there a way to fix it (even increasing the number of decimals nothing happens)?
Thank you

Luigi
Attachments
std.jpg
(68.17 KiB) Not downloaded yet
admin
Site Admin
Posts: 822
Joined: Mon Aug 13, 2007 11:48 am
Location: ISTerre
Contact:

Re: anomalous Standard Deviation

Post by admin »

Hi Luigi,

Try to set it manually to 1.1. Is it better?
Best regards,

Marc

PS: your message went out of the list of messages to approve by mistake. During Christmas holidays I did not connect to geopsy.org.
luigiV
Posts: 121
Joined: Fri Jul 17, 2020 9:28 am

Re: anomalous Standard Deviation

Post by luigiV »

Hi Marc,
with 1.1 is fine. Till present I had always successfully used 0.1.
Is there a reason for this?
Thanks
Luigi
admin
Site Admin
Posts: 822
Joined: Mon Aug 13, 2007 11:48 am
Location: ISTerre
Contact:

Re: anomalous Standard Deviation

Post by admin »

Was it with 3.4.2? With older releases the uncertainty where not processed the same way and it is possible that 0.1 was fine.
luigiV
Posts: 121
Joined: Fri Jul 17, 2020 9:28 am

Re: anomalous Standard Deviation

Post by luigiV »

Of course,3.4.2, that's why I was surprised both that 0.1, which always worked, was no longer correct, and that 1.1 gave the right result, overnight.

With Geopsy 2.10 to get 10% I have to write 0.001.

Regard

Luigi
admin
Site Admin
Posts: 822
Joined: Mon Aug 13, 2007 11:48 am
Location: ISTerre
Contact:

Re: anomalous Standard Deviation

Post by admin »

I don't why either. If you manage to make it work with 0.1 again, try to identify the context and conditions. If I can also reproduce it, it will be easier to fix.
luigiV
Posts: 121
Joined: Fri Jul 17, 2020 9:28 am

Re: anomalous Standard Deviation

Post by luigiV »

I have tested previous dispersion curves that correctly represented 10% with 0.1 and found that they all now have the anomaly of representing 10% with 1.1.
Previously I had recorded that the mean curve automatically provided absurd standard deviation bars, equal to the ones I now obtain by applying an SD of 0.1 (the two curves to be averaged have no SD, see attached jpg).
If I apply an SD of 10% with 1.1 to the two curves, I obtain an average curve with an SD that is >30%. Is it wrong or am I wrong (see the jpg STD_average_1)?
Attachments
std_average.jpg
(65.62 KiB) Not downloaded yet
std_average_1.jpg
(54.72 KiB) Not downloaded yet
Post Reply