Well, I think I'm finally banging up agains the 10 station limit incorporated into the Demo version of Star*net, I now realize it's THE must-have piece of software going forward, but unfortunately, it's just not in the budget. So....
As a work around, Is there a way to split a Starnet file into segments, keeping one or two points common to the resulting two files, without losing the value of Starnet? I believe there's a way to manually assign uncertainty (or the std errs) for individual stations. If I ran an analysis on file "A", found the uncertainty at the "far end" stations, then brought them into a new traverse segment, would that work?
I know it's a kludge, but necessity is the mother of invention for me right now.
As a side note: If any one has a seat they'd like to sell (affordably), email me please.:-)
> As a work around, Is there a way to split a Starnet file into segments, keeping one or two points common to the resulting two files, without losing the value of Starnet? I believe there's a way to manually assign uncertainty (or the std errs) for individual stations. If I ran an analysis on file "A", found the uncertainty at the "far end" stations, then brought them into a new traverse segment, would that work?
If you had azimuth observations on each segment with uncertainty estimates for them, you could split the traverse up into segments, compute the inverse between the endpoints together with the uncertainties in length and bearing and create a synthetic traverse consisting of just those inverses (assuming they are all connected.
You'd adjust that figure and then use the adjusted coordinates of the endpoints to control the adjustment of the individual segments. You'd lose the ability to calculate relative uncertainties across the survey, aside from between points on the larger synthetic traverse, but the adjusted coordinates ought to be quite close to what the fully functioning version of Star*Net would generate by running the whole survey as a single adjustment.
> If you had azimuth observations on each segment with uncertainty estimates for them, you could split the traverse up into segments, compute the inverse between the endpoints together with the uncertainties in length and bearing and create a synthetic traverse consisting of just those inverses (assuming they are all connected.
Come to think of it, you wouldn't necessarily need separate azimuth control for each segment of the traverse as long as you had a reference azimuth of known uncertainty at the beginning of the traverse.
Starting from there, you'd enter as much of the traverse as you could handle in the demo version of Star*Net and then run that through Star*Net to estimate the coordinates and uncertainties of the next to last point on the traverse segment and the uncertainty in the bearing of the last leg of the segment.
Then, to compute the next segment, you'd begin at the coordinates of the next to last point with its uncertainties, and enter the length and bearing of the last leg of the previous segment (with the uncertainties that Star*Net estimated) as the first leg of the next segment, and so on until the end of the traverse.
The overall adjustment would take the form of breaking the figure down into longer resultant courses representing each segment and with the uncertainties Star*Net computed for them.
The input files would have to be carefully constructed to allow the inevitable multiple runs that would probably be required. It would be fairly labor intensive, but an option. What would be even easier would be to learn how to post the ascii data of the input file here and just prevail upon someone to run the adjustment and send the results to you.
> The input files would have to be carefully constructed to allow the inevitable multiple runs that would probably be required. It would be fairly labor intensive, but an option. What would be even easier would be to learn how to post the ascii data of the input file here and just prevail upon someone to run the adjustment and send the results to you.
Thanks, Kent. Reading the detailed instructions and things to consider lead me to believe this is a fools errand. Given my lack of experience with Starnet, I must ask myself: Do these additional things to consider help, or get in the way of, learning to craft input files that will result in the best analysis of a small network, and from there, to learn how to add measurements that will add to the accuracy?
The answer to that is: craft the input files and post them. Not only might I get the results of the analysis, I may get valuable insight (not to mention error correction) which is often offered up here.
I'll go that route until I own the program.
> The answer to that is: craft the input files and post them. Not only might I get the results of the analysis, I may get valuable insight (not to mention error correction) which is often offered up here.
>
> I'll go that route until I own the program.
One useful tip for posting formated ascii data is to use the [_pre] and [_/pre] tags before and after it, (but without the underscore character).
>
> One useful tip for posting formated ascii data is to use the [_pre] and [_/pre] tags before and after it, (but without the underscore character).
You mean like this?:
[ pre]
This is a test.
B 1-14 s31-02-39w !
B 14-13 s58-27-21w
#B 2-1 s56-23-13e
D 1-14 1068.22
D 1-2 230.8
D 2-3 329.545
D 2-3 329.565
[ /pre]
I don't think it's working; The tags should not show in preview, correct?
> You mean like this?:
> [ pre]
> This is a test.
> B 1-14 s31-02-39w !
> B 14-13 s58-27-21w
> #B 2-1 s56-23-13e
>
> D 1-14 1068.22
> D 1-2 230.8
> D 2-3 329.545
> D 2-3 329.565
> [ /pre]
>
> I don't think it's working; The tags should not show in preview, correct?
You have an extra space in the tags that shouldn't be there. This is what the format looks like. It's in monopitch courier type which means that columns that line up in the ascii will as posted :
[pre]
This is a test.
B 1-14 s31-02-39w !
B 14-13 s58-27-21w
#B 2-1 s56-23-13e
D 1-14 1068.22
D 1-2 230.8
D 2-3 329.545
D 2-3 329.565
[/pre]
Got it. Thanks.
[pre]
This is a test.
B 1-14 s31-02-39w !
B 14-13 s58-27-21w
#B 2-1 s56-23-13e
D 1-14 1068.22
D 1-2 230.8
D 2-3 329.545
D 2-3 329.565
[/pre]
Well, I'm skeptical as to whether this will work. I'm still learning the proper data file formats, and have been running blunder check continuously, but you can't even do that with the demo version. This will be like flying blind.
At any rate, I went back to an traverse I tried this summer, but started from a different point and got back to my original POB.
If anyone is up for running the data through Starnet, I'd appreciate it. If it flies, I have a few questions about a few of the points.
Thanks,
[pre]
C 1 1000 1000 ! !
B 1-2 n56-23-13w !
B 1-14 s31-02-39w !
B 14-13 s58-27-21w
D 1-14 1068.22
D 1-2 230.8
D 2-3 329.545
D 2-3 329.565
D 2-3 329.505
D 2-9 376.37
D 2-9 376.365
D 2-9 376.365
D 3-4 170.195
D 4-5 175.12
D 5-6 271.084
D 8-9 384.65
D 9-10 108.770
D 6-8 122.145
D 14-13 150
D 10-2 292.575
D 10-3 140.46
D 31-32 513.867
D 32-31 513.690
D 32-33 24.937
D 28-30 245.005
D 29-30 194.062
D 29-21 345.270
D 23-25 259.122
D 25-34 450.015
D 25-23 259.105
D 34-35 465.700
D 34-25 450.020
D 34-7a 87.275
D 35-1 352.950
A 2-10-3 92-17-57
A 3-2-9 37-07-04
A 3-10-9 133-30-33
A 1-2-3 94-57-47
A 1-2-3 94-57-30
A 1-2-3 94-57-43
A 1-2-9 132-05-13
A 1-2-9 132-05-10
A 2-3-4 177-08-18.5
A 3-4-5 142-58-33
A 4-5-6 282-08-17
A 5-6-8 245-48-55
A 6-8-9 229-58-53
A 8-9-2 218-54-50
A 9-10-2 134-12-07
A 9-2-3 322-52-27
A 30-31-32 192-37-56
A 31-32-33 159-45-01
A 21-28-30 268-25-41
A 21-29-30 262-13-15
A 14-23-25 275-36-11.5
A 23-25-34 182-24-16
A 25-34-35 229-30-05
A 25-34-7a 291-14-02
A 34-35-2 223-16-02
A 34-35-36 282-22-25
[/pre]
I've run it, I'm getting errors - no solution.
Is your angle AtFromTo or FromAtTo?
There is no distance to 36.
> I've run it, I'm getting errors - no solution.
>
> Is your angle AtFromTo or FromAtTo?
>
> There is no distance to 36.
From-At-To
I believe it's D 35-36 174.110
I pasted this into my home-grown program, which is a pale imitation of part of Star*Net with slightly different syntax.
My program doesn't (yet) have centering errors automated - I have to manually calculate those into each angle or distance standard error, and I didn't take time to do that on this example, so the statistics aren't particularly meaningful.
The main thing that jumps out at me is that there is a group of points that don't seem to be tied down to the rest of the points at all, or to each other very well:
21 28 29 30 31 32 33
I can almost arbitrarily fix some coordinates a large distance from where they fell, or add distances, or angles without greatly increasing the overall error of fit. For instance I fixed these, with point 30 easting at 1500 and some other made-up values:
21 northing
30 easting and northing
33 northing
Are there some more angle and distance measurements to include, or did I lose something in the transfer?
> Are there some more angle and distance measurements to include, or did I lose something in the transfer?
Probably the former.:-O
I'll draw a sketch of what it should look like and carefully check that I have each angle and distance required. Thanks all so far for the input.
A more careful look shows that 21 28 29 30 are tied with sufficient angles and distances to each other, but that group is only tied by angle to 31 and hence the 31 32 33 subgroup. Neither subgroup is tied to the rest of the points in any way.
The uncertainty ellipse at 23 is large, and would be greatly improved by having a distance 14-23. If there is a line of sight to get an angle to 14, that distance should be measured.
What standard errors are you using for centering, distance, and angle?
> A more careful look shows that 21 28 29 30 are tied with sufficient angles and distances to each other, but that group is only tied by angle to 31 and hence the 31 32 33 subgroup. Neither subgroup is tied to the rest of the points in any way.
>
> The uncertainty ellipse at 23 is large, and would be greatly improved by having a distance 14-23. If there is a line of sight to get an angle to 14, that distance should be measured.
>
> What standard errors are you using for centering, distance, and angle?
Here's the sketch.
First, to answer the questions:
D 23-14 167.937
I'm missing other angles/distances from the original traverse loop (2-3-4-5-6-8-9). I'm trying to dig them up and will update the file.
I'm using 5" for Directions, 6.8" for angles. (I think that's what the final setting was for my new machine that produced a near 1.0 error factor for directions...I'll check my other thread (TS Accuracy Chapter XX).
Centering errors .005 is probably in the neighborhood.
Distance: +/-2mm plus 2ppm.
My intention is to hold 1-2, 1-14 and the 14-13 backsight fixed...They're from a known good, recent survey.
28 and 33 are corners of the property; I'd love to know what the traverse ends up calling the call between 33 and 2. I don't have the software to do this (other than a lot of trig and my HP11c).
My next move will be to traverse from 33 to 34, but it's very steep, all wooded with a gazillion sticker brambles, it's getting really cold out and the snow is about to fly, so I may not get it done soon. (I know...complain complain complain)
The sketch helps, even though it is not to scale. I note that it is missing points 10 and 36 for which you have measurements (is 36 same as another point for checking closure?). And of course it has several points for which you had not posted data.
Point 13 doesn't affect anything else so it is unimportant, but the sketch does not match the bearing given for 14-13.
Are you sure about D 14-23 ? I can't get the 2-35-34-25-23-14 loop to fit at all with the new distance 14-23. It is pulled to the east 2-9-8-6 etc, whereas your sketch shows it to the west.
> The sketch helps, even though it is not to scale. I note that it is missing points 10 and 36 for which you have measurements (is 36 same as another point for checking closure?). And of course it has several points for which you had not posted data.
Mistakes Mistake Mistakes...ADD on the loose again!
D 34-36 174.110 It is indeed a point from an earlier traverse I didn't note on the drawing. It should be very close to a line drawn between 2 and 33.
>
> Point 13 doesn't affect anything else so it is unimportant, but the sketch does not match the bearing given for 14-13.
The bearing TO 13 from 14 is S58-57-21E. NOT W
>
> Are you sure about D 14-23 ? I can't get the 2-35-34-25-23-14 loop to fit at all with the new distance 14-23. It is pulled to the east 2-9-8-6 etc, whereas your sketch shows it to the west.
Not sure I understand, unless the bearing mistake from 14-13 screwed things up. I have one more recent (and I feel really good measurement from 14-23: 167.745 (compared to 167.947), but I don't think that's the difference you're talking about. Something else must be wrong.
I'm re-checking all my distances and angles again. Keep in mind I'm not 100% electronic yet. I'm still reducing and copying numbers from the field book into a spreadsheet (or Starnet). That's part of my problem right there.
Here's the revised file with the missing measurements (the ones I could find so far at least):
[pre]
C 1 1000 1000 ! !
B 1-2 n56-23-13w !
B 1-14 s31-02-39w !
B 14-13 s58-27-21e
D 1-14 1068.22
D 1-2 230.8
D 2-3 329.545
D 2-3 329.565
D 2-3 329.505
D 3-4 170.195
D 4-5 175.12
D 5-6 271.084
D 6-8 122.145
D 8-9 384.65
D 9-10 108.770
D 10-2 292.575
D 10-3 140.46
D 14-13 150
D 14-23 167.745
D 23-25 259.122
D 23-26 242.6
D 23-24 489.685
D 25-23 259.105
D 25-24 230.596
D 25-34 450.015
D 24-6 114.143
D 31-32 513.867
D 32-31 513.690
D 32-33 24.937
D 28-30 245.005
D 29-30 194.062
D 29-21 345.270
D 34-35 465.700
D 34-25 450.020
D 34-7a 87.275
D 35-1 352.950
A 2-10-3 92-17-57
A 3-2-9 37-07-04
A 3-10-9 133-30-33
A 1-2-3 94-57-47
A 1-2-3 94-57-30
A 1-2-3 94-57-43
A 1-2-9 132-05-13
A 1-2-9 132-05-10
A 2-3-4 177-08-18.5
A 3-4-5 142-58-33
A 4-5-6 282-08-17
A 5-6-8 245-48-55
A 6-8-9 229-58-53
A 8-9-2 218-54-50
A 9-10-2 134-12-07
A 9-2-3 322-52-27
A 30-31-32 192-37-56
A 31-32-33 159-45-01
A 21-28-30 268-25-41
A 21-29-30 262-13-15
A 14-23-25 275-36-11.5
A 23-25-34 182-24-16
A 25-34-35 229-30-05
A 25-34-7a 291-14-02
A 34-35-2 223-16-02
A 34-35-36 282-22-25
[/pre]
I appreciate the help. I can see the writing on the wall though...I've got to get myself Starnet.