Semi Variogram Paramaters Window

Hi Hans,
I use SAGE to model my variograms and I always enter them to Micromine as you are doing.  I bypass the semi-variogram section completely under the STATS menu.  It is very fussy and I would say, odd.  Instead, I recommend that you click on Search Ellipse in the Vizex Forms window, set up a new Data Search.   Open the new search tab, Ellipsoid Properties, and enter your Supervisor directions in that dialog.  I usually use the geostatistical option and try entering the principal axis first, then on the second line (2), the plunge of the intermediate axis.  Remember that Micromine dip signs may be opposite your software, the case with SAGE.  With the inputs of the principal axis azimuth and plunge, and the intermediate plunge, the rest of the dialog populates.  Occasionally, I need to enter a different axis on line 1 to get a good match to SAGE, but it usually one takes a minute or two for the whole process.  Now, screen capture the calculated azimuths and dips on line 2 and line 3 and enter them in your variogram parameters profile in the Ordinary Kriging dialog with the stated precision.  You will not ever get an error doing it this way.  Micromine is very fussy about precision but at least the variogram parameter dialog and the search ellipse dialog values are synchronized.   Good luck!

Agree with both of you that the current setup is painful and that mutual orthogonality error message can drive one absolutely insane. After chatting with the Micromine crew, I was under the impression that variography was due for an serious update; however, while the current 2014 beta version improves the descriptive statistics functionality, there's no sign of change to variography. 

Like Hans, I've taken Micromine out of the equation for model fitting and production of plots for reports. I use Mathematica and have found it to be exceptionally capable once you get over the language hurdle. I find geostatistics and spatial analysis makes far more sense as a flowing workbook complete with sections, annotations and clean visualisations, rather than a series of disjointed individual functions that have to be compiled into Word. 

An option for you may be to investigate the python scripting functionality available from MM2013 onwards. I compile and store critical variogram model parameters in a *.csv file within the project. This parameter file is convenient as you can edit it directly in Excel or output an entirely new set your chosen variogram modelling platform. I'd imagine that Supervisor would be able to export something suitable. 

The file is imported via Python scripting and the parameters are automatically split into the various fields (directions, nugget, sill, range, model type etc.) of the estimation model dialogs we all know only too well. The trick is that you create the model parameter forms entirely via Python, so you never need save a variogram model as a formset again. All it takes is a means to reference the correct parameters back to their corresponding domains within the estimation. 

One potential advantage of this approach is that the orthogonality validation event seems to be triggered by hitting the OK button on the directions form and it's likely that creating the form in code won't trigger the event and may pass the parameters in their specified state. I haven't checked this aspect yet and don't know how it would behave, but it's not likely to jeopardise the estimation if we're talking a minor precision error. It's just a question of whether the Kriging algorithm would accept the 'invalid' parameters. As Hans mentioned, it could well be that your directions make complete sense geometrically, but Micromine simply expects the opposite of what you're entering.

It would be handy if Micromine would provide the programming logic behind the automatic validation. If we knew how it goes about validation we could develop the a Python function to allow us to ensure compatible directions every time. 

Anyhow, keep at it and let us know if you get stuck.

Olly, Don and Hans,

Yes, the variography is undergoing a major update for Micromine 2014. We are working on it now; it was simply not mature enough to include in the Beta release.

Our main focus is on creating an effective workflow, and to that end we have added a new variogram map option to the menu, which draws the map using polar-coordinate sectors (in much the same style as Isatis). We are also modernising the variogram modelling form and will add options for handing data between the variogram map and variogram graphs without having to re-enter it. From there it is only a small change to add support for third-party variogram models.

Given your respective workflows we would be very grateful for suggestions for (and examples of) the types of data you think should be exchanged. Please do send them in as now is the perfect time to guide the direction we take with our new workflow. If you have my contact details please feel free to contact me directly; otherwise feel free to use the Contact Support option in the Help menu.

For now the simplest way to automate the importing of variogram model data into Micromine would be to write a Python script to do it, as per Olly's suggestion.

Regarding the orthogonality requirement, I appreciate the frustration of having to enter the angles to several decimals of precision before they are treated as valid. We made this decision because we were getting more support calls and bug reports back when Micromine used to visibly round the numbers. With the new workflow this won't be an issue as Micromine will simply pass the full-precision values from one place to the next.

Olly, regarding your question on the orthogonality validation, it simply checks the dot-products of the three 3D vectors and ensures that they are within around 0.5° of being at 90° to each other (the true tolerance is based on the arc-cosine of the dot product). There was a brief period when the tolerance was much less than that, which was fixed early last year, so please do ensure you're using the latest Micromine 2013 service pack.

Regards,

Frank

Here is a preview of the new variogram map. So far we have only completed the map display; we still need to add the directional selection (and potentially the measurement of an anisotropy ratio):




What might not be obvious in the image is that clicking a cell in the map has selected the corresponding samples in Vizex.

Most of our new statistics tools work this way once you enable the Sync Selection options from the corresponding window or toolbar: if the same file is loaded in Vizex, the File Editor, and two different Stats windows, making a selection in any one of them will select the corresponding records in the others.

It's a great way to see how many lags are affected by any individual sample, or to locate potential outliers by selecting them in the histogram or box plot.

Looks good, will be great to see the completed version.

Thanks Ron. I'm sure you have your own thoughts on this so please do feel free to add your suggestions to the mix.

Guessing Sync Selection is rolling out in a later beta Frank? Current beta seems to have the buttons in the graph windows, but the option in Vizex is currently hidden? Seems a really good idea.

On the variography, it would be great to be able to compile all your directions in a single window. I find the single plot per window isn't a very effective use of space and tiling the windows vertically/horizontally isn't great. I tend to work with a host of smaller plots similar in style to a matrix plot to provide a good overview for comparative work. 

Also the ability to change the scale of the axes would go a very long way. Auto scaling can be a right pain when you're looking at small quantities and the variance spikes.

Would this be better in a dedicated thread?

Hi Frank,

I would say linking of views, variogram maps and tables but looks like your onto that. From a variography point of view it would be great if we could do correlograms and gaussian variograms (ie do a Gaussian/normal score transform of the data - model the variogram on this and then do a back transform of the variogram parameters to use them in an OK estimate). I find if I can not use raw I will fall back to a correlogram or Gaussian and at the moment I have to go outside MM to do that. The tidy up of the forms and process you seem to be doing is is a big plus.

@Olly, yes you're right about going off-topic. I've created a new thread specifically for the discussion of new variography features. You can get to it at: http://forum.micromine.com/topic/414388-suggestion...

I'll move the existing posts over to the new thread (once I figure out how to do it!)