[Therion] Therion does not write valid colour by horizontal or vertical error data to 3d output

[Bruce Mutton]

Hi Benedict

If ‘loop-closure survex’ is set in therion.ini then Therion should extract the horizontal and vertical error data from \thTMPDIR\data.3d.  Survex calculates these as the number of standard deviations from expected error, or ‘sigmas’.  Ratio of actual errors over expected errors (ie weighted by survey grade or readings sd).  Olly may like to clarify.

 

If ‘loop-closure therion’ is set in therion.ini the Therion should, as well as calculating the absolute ‘misclosure / length of loop’, also calculate ‘horz component misclose / length of horz components of loop‘ (or something along those lines), and populate the .3d output accordingly.

 

Olly explained it quite well somewhere, which I have not found just now, but it you trawl up and down this thread

https://www.mail-archive.com/therion@speleo.sk/msg06283.html

you will find some discussion of the concepts.

 

The bug (or feature request) that I am reporting is that Therion makes no attempt to export the horz and vert error component data to .3d output files.

 

As an aside, Therion and Survex calculate the error data quite differently (as indicated above).  This results in quite dramatically different perceptions of how good or bad a survey is.  I think they are both useful and I am not suggesting that either be changed, but in general I prefer and use the Survex method.

Here is one comparison showing colour by (total) error for the same cave (total because that is all that Therion will currently export to 3d correctly).

With ‘loop-closure therion’ the loops are generally between 1% and 4% with one small one at bottom left about 8% or 9% and one near top right at 7%.



 

With ‘loop-closure survex’, there are about three loops at 12 sigma and not so many in the preferred range of 0 to 3 sigma.



 

Comparing the two you can perhaps see that Therion’s method tends to make longer loops look better than perhaps they really are.

A nice feature of Therion’s method is that it is independent of survey grade, whereas Survex’s method makes low quality loops look ‘blue’ (good) if they are assigned a low survey grade (this is the intention of course, but it would be nice to be able to toggle out the effect of the survey grades from time to time – maybe there is scope for a ‘colour by grade’ feature in Aven?).

 

If and when we can separate out the errors by horizontal or vertical component, we will more easily be able to localise where the poor sections of survey are.

 

Bruce

 

From: Therion <therion-bounces at speleo.sk> On Behalf Of Benedikt Hallinger
Sent: Sunday, 26 April 2020 21:44
To: List for Therion users <therion at speleo.sk>
Subject: Re: [Therion] Therion does not write valid colour by horizontal or vertical error data to 3d output

 

What is the expected data therion should write instead?





Am 26.04.2020 um 07:00 schrieb Bruce Mutton <bruce at tomo.co.nz <mailto:bruce at tomo.co.nz> >:



A while back Survex 1.43 improved the ability of Aven to visualise loop misclosures.

Not only can it colour by error, it can now colour by horizontal or vertical loop error.

 

Unfortunately, the 3d files produced by Therion do not properly write the data.

A brief discussion here https://github.com/ojwb/survex/commit/fe8a1b478b91bd977a4b21be49af25724b16c692 that identifies the problem.

Olly suggests the problem is:

-Therion writes 0 for all the horz and vert error values where Survex is being used for data processing.

-Therion equates the horz and vertical error to the total error where Therion does its own data processing.

 

For users who want a work around, there is one described in the last line of the thread linked above.

 

Bruce

 

_______________________________________________
Therion mailing list
Therion at speleo.sk <mailto:Therion at speleo.sk> 
https://mailman.speleo.sk/listinfo/therion

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.speleo.sk/pipermail/therion/attachments/20200427/27bd4ae9/attachment.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image002.png
Type: image/png
Size: 79037 bytes
Desc: not available
URL: <http://mailman.speleo.sk/pipermail/therion/attachments/20200427/27bd4ae9/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image005.png
Type: image/png
Size: 78050 bytes
Desc: not available
URL: <http://mailman.speleo.sk/pipermail/therion/attachments/20200427/27bd4ae9/attachment-0001.png>