Measurement groups that contain values out of tolerance

Hi Micha,
I have already written on this forum for other questions, always receiving clear and precise answers.

We use your program in one of the most important projects in northern Italy, with considerable satisfaction and I hope this will please you.

But, let's go to the question: we use a very important amount of data in the calculations entrusted to your software, so we have many groups relating to the horizontal angles, distances and Azimuth that we still keep divided by campaigns. We find it a bit difficult to report any value that does not satisfy the various tests to the belonging group.
Would it be possible to have a column in the Raw and / or Result Data table indicating the group from which the measure comes?
You can also think of other solutions, eg. highlight with a color the "item"group that contains the measurement that did not pass the tests.
I leave it to you to find an easy way to find a possible solution.

Thank you in advance for the help you can give us.

Amedeo

Hello Amedeo,

We find it a bit difficult to report any value that does not satisfy the various tests to the belonging group.
Would it be possible to have a column in the Raw and / or Result Data table indicating the group from which the measure comes?

I understand the problem but not how you will use this information in the next step. If you select more than one group in the tree menu, the date of these (selected) groups are shown in the table. If you sort the table by e.g. the test statistic Tprio, it is (nearly) impossible to identify the group containing a specific observation. That is the behaviour we are talking about, right? My question is now: What do you do with such an information? Are outliers treated differently, if they come from different groups?

You can also think of other solutions, eg. highlight with a color the "item"group that contains the measurement that did not pass the tests.

Currently, you can highlight the rows of the table (not the tree item), but this does not give you the information about the group an observation comes from. Please note, the tree item is not the group itself, it just represents the group. The data of the group are loaded on request (in the moment you click on the item).

[image]

I leave it to you to find an easy way to find a possible solution.

It would be great, if you can give me some information about my question. If I understand your problem better, it is easier to find a solution.

Kind regards
Micha

--
applied-geodesy.org - OpenSource Least-Squares Adjustment Software for Geodetic Sciences

Thanks for the reply, Micha.

I know the procedures you showed me, but once you have identified the values that have not passed the tests, in a project that contains 4000 angular measurements divided into 100 campaigns (Groups), analyzing the stationing in which the anomalous value resides, is not difficult but rather slow, because you have to scroll one by one or in groups to find the stationing with values out of tolerance.

We have organized the work in Jag3d by measurement campaigns and therefore by groups, each campaign is named with a sequential number, the date, and obviously the station to be oriented.

What do I do with it then?

Nothing particular, once I have the value out of tolerance and the corresponding campaign (group) I go to recheck the data I have loaded and check the anomalous angular values with respect to the history that I have available on other software and I make my decisions about it.

I believe it is an implementation that would greatly facilitate the analysis and "debugging" of the uploaded data.

If you think it is appropriate I can send you a sample project.

Grazie.
Amedeo

Hello Amedeo,

What do I do with it then?

Nothing particular, once I have the value out of tolerance and the corresponding campaign (group) I go to recheck the data I have loaded and check the anomalous angular values with respect to the history that I have available on other software and I make my decisions about it.

Okay. So the next step is to identify the group(s) containing the observations and to make some kind of analysis. I have thought about this procedure and I've looked for a solution which is more generic and not restricted to a specific task e.g. outliers. I've added a new item "Select item groups" in the context menu of the table. The following screenshot depicts the new option. It is some kind of backward search and selection. It supports multi-selection of the table rows.

I've uploaded a modified JAG3D version that contains this feature. Please take look at this version and let me know, how it works and if it is helpful during your analysis.

Download: JAG3Dv20210522_Windows.zip

(I only upload the Windows version. Let me know, if you are working on Linux or MAC)

Kind regards
Micha

--
applied-geodesy.org - OpenSource Least-Squares Adjustment Software for Geodetic Sciences

Tags:
JAG3D, Group, Selection, Context menu, Table

Hi Micha,
Simply perfect.
I have tried in both the angle and distance groups and it works fine.
This makes our work much easier.
Thank you
Amedeo

Hello Amedeo,

Simply perfect.

Thank you for your reply and your positive feedback. I'll provide this feature within the next release.

Kind regards
Micha

--
applied-geodesy.org - OpenSource Least-Squares Adjustment Software for Geodetic Sciences

Dear Amedeo,

the latest JAG3D version contains your suggested feature on group re-selection by data. Please finde this version here: JAG3Dv20210606.

Kind regards
Micha

--
applied-geodesy.org - OpenSource Least-Squares Adjustment Software for Geodetic Sciences

Tags:
JAG3D, v20210606

RSS Feed of thread