All this talk of least squares has got me looking at SurvNet again. However, I am struggling with it processing my raw data files. Normally, I use Carlson's regular Raw Data editor routines. The several times I have tried Survnet in the past, I struck out on the point numbering. Now that I have discovered that I can add an "=" in my descriptions for my closing point, I've been able to move forward.
The same file that will work in Carlson's Raw Edit Least Squares gives me problems in SurvNet. I get a "Math Overflow from iterations" right off the bat. I've tried specifying the traverse closure file, but I don't think that affects the adjustment. Other times I will get "Zero Degrees of Freedom".
So far I have been able to get one file to converge in 2 iterations, but I have not been able to reproduce the results.
One question- Do I need to reduce the direct-reverse sets in the raw data editor before bringing the RW5 into SurvNet?
Overall, I think Survnet is reading the RW5 wrong and it cannot converge because certain data is in left field.
Any tips on RW5 formatting (i.e. SS vs. TR usage) to get this thing to work?
I gotta walk away for a minute- this thing is giving me a headache.
> One question- Do I need to reduce the direct-reverse sets in the raw data editor before bringing the RW5 into SurvNet?
First - my experience with SurvNet is very limited. As a matter of fact, I used it once, maybe twice, and saw that it had a lot of similarity to StarNet. So my ability to help is limited.
That said, I'm certain that your direct/reverse sets should be reduced to a single angle for each set and that the import process should be doing that reduction for you.
>
> One question- Do I need to reduce the direct-reverse sets in the raw data editor before bringing the RW5 into SurvNet? No You do not have to reduce the data. The raw data just the way you turn it in the field will work
>
> Overall, I think Survnet is reading the RW5 wrong and it cannot converge because certain data is in left field. Check the begining of the report. SurvNet lists any observation that are more than the tolerance settings. This is where I look first. One of the most common issues is having the same point number for two different points
>
> Any tips on RW5 formatting (i.e. SS vs. TR usage) to get this thing to work? No real secret, the software will use the observations in most any order.
The Carlson Web Page has a webinar on the subject at
http://www.carlsonsw.com/support/carlson-movies/#Survey . This is definitely worth watching.
I've seen the webinar Dean Goodman put on. Problem is, he used a C & G file (and probably prepped data) for the demo and I'm trying to use an RW5 with raw data from a SurvCE data collector.
The file I am working with at the moment gives me an error saying it is missing an angle from "bs:2 ip:3 fs:21". That data is in the file. I am having to chop the RW5 file up in an attempt to figure out why the process won't pick that info up. Like I said earlier, Carlson's native Raw File Editor processes a Compass Rule / Least Squares, no problem. It's a four station traverse with D&R at each station. It does not like the D&R data, so I have the Raw Data editor reduce that to the meaned angles. Then SurvNet says I have zero degrees of freedom....
If it does not process you may have a couple problems.
1. look at your back-sights make sure you have erased all the bad ones or extra entrys.
2. make sure your = are on the right points. Make sure you have not used numbers twice for two separate shots.
3. Did you give it at least a minimum of two control points (NE) in the top of the raw file.
4. Did you set CSE in the top of the file .
5. leave your doubles alone they are fine.
Send me a copy of the raw file I will look it for you .
I would assume that the zero degrees of freedom statement from SurvNET is telling you that you have the minimal amount of observations to determine the figure you are attempting to adjust. Do you have any redundant measurements you are not including? You may need to turn another angle or two to get that message to change.
We use RW5 files as well. I suggest that you call the Carlson Support Desk. Give them a shot and if they can not help then tell them you want to speak to Dean Goodman. Dean has looked at more than one of our raw files and solved the problem. Learning the ins and outs of this software is well worth the investment.
I agree, I definitely want to take full advantage of the software. Glad to know they are willing to work with us.
Haven't messed with SurvNet in a while, but I remember (perhaps incorrectly) that it could not be formatted like the typical rw5 in that you DO have to use the same point number, not a different # with the = in the description. I recall being irritated that I had to have 'two versions' of the same file.
Not true . Same raw file and "=" works fine with different numbers . I use Survnet everyday its works good for us .
I sat in on a Survnet seminar last February at the KAPS Conference. If I remember correctly, you can have point 95 be your closing point, with the description being =1. The description, as I understand it, is telling the software that point 95 is your closing point, and should have the same ending coordinates as your starting point.
I would call Carlson Support. Mr. Goodman was pretty helpful at the seminar, and said he would welcome calls from users with questions.
Mark data Ran fine
Marks data ran fine in Survnet 
Reading this: http://files.carlsonsw.com/mirror/manuals/Carlson_2015/source/Survey/Survey/SurvNET/Introduction/Introduction.html W/F/S
I found the following: Some data collectors do not allow the surveyor to shoot the same point twice using the same point number. SurvNet requires that all measurements to the same point use a single point number. The raw data may need to be edited after it has been downloaded to the office computer to insure that points are numbered correctly. An alternative to renumbering the points in the raw data file is to use the 'Pt Number substitution string' feature in the project 'Settings' screen. See the 'Redundant Measurement' section for more details on this feature.
I'm not looking at SurvNet, but I guess the reference to 'Pt # Substitution String' above may be what allows "=" to be used. I would guess that had to be set up and some button checked somewhere; the lack thereof might be causing OP's issues.
Yes "=" in the description allows you to use any number in the shot . So "=1" for shot 100 will tell the program its the same as number 1.
Hi Mark,
I would be happy to help you with your SurvNET projects. You can contact me at dgoodman@carlsonsw.com. If you have trouble processing a RW5 file send it to me and I will take a look at it.
SurvNET is much more forgiving than the traditional Carlson adjustments. The biggest problems clients have at first is making sure the point numbers are correct and the control is defined properly. Once those are addressed everything will work.
There are tools that allow you to track down point numbering problems. Use the Process->Blunder Detection->Preprocess Raw Data option. The K-Matrix test will show up point numbering issues. The "Proximity" report will show points that may be the same point.
Dean Goodman
Thanks Dean. Djames was able to point me in the right direction. You are correct- I did not have the control points specified properly. I started with an assumed coordinate of 5000,5000 for my first point and had an azimuth to the backsight point. Once I entered the coordinates for the backsight point and added the CSE data to the top of the RW5 file with an exclamation point, it worked.
I will keep you in mind for the next closed traverse I run.
Also, do you have any "superuser" tricks/tips for SurvNet you would like to share with the board?
