Translating or Rotating an excel list of points to fit the coords of a benchmark

1555
7
Jump to solution
01-27-2013 05:04 PM
KenLucas
Occasional Contributor
I have an excel list of the coordinates created for a list of points. The last point in the list was to a survey monument, for which I have the original, accurate coordinates. I'd like to adjust the coordinates for all of the points to fit the actual coords of the survey monument. I have an ArcInfo 9.3 license, with the COGO toolbar and spatial analyst. I don't have Survey Analyst. The cogo toolbar seems to only allow adjustment of traverses, not collections of points. I'd like to solicit comments from the community about the following ideas:
If I knew the amount of direction & distance I wanted all of the points to move (to match the adjustment of the last point to the monument) should/could I adjust the coordiates in excel before creating a point feature? 
Could I select all of the points in arcmap and globally shift them by a specified direction & distance? I realize I could simply just select all of the points and shift them, with the editor tool, but I can't figure out how to, globally, shift the points a specified distance and direction. The edititor direction - distance tool does not shift all of the selected points at once.
I've heard of a toolbar called the Parcel Edititor; would this tool accomplish my objective? If so, what gis extension is the parcel editor associated with? 
Ken
Tags (2)
0 Kudos
1 Solution

Accepted Solutions
T__WayneWhitley
Frequent Contributor
Ken, I'm not saying this is the way to go to ultimately solve your problem, but if you want to try another shot at Feature to Point at least to have a look at it, you must 1st have true multipoint input to feed into it, i.e. a point 'group' per feature record.  How to get this?  Use of the Dissolve tool is one way, using a common ID by which to group.  Or if you don't have that ID coded right away and just want see what I mean, then select in a map a small cluster of points to experiment with (they can be randomly selected for purposes of this experiment), open Dissolve and enter the point feature layer you have a selection on and run Dissolve (without specifying a field to dissolve by)....check the output- open the attribute table, and it should be a single record.  Selecting the record will highlight the whole group (the points are now 'parts' of a different kind of 'whole', called multipoint).  Input that into Feature to Point and it should give you the centroid of the multipoint geometry.

Hope that's clear.


Enjoy,
Wayne

View solution in original post

0 Kudos
7 Replies
PhilipMorrison
New Contributor
Ken,

I actually have no idea how you would do what you are suggesting.  I am still a relative rookie when it comes to ArcMap.  However, my comment relates to why you would want to do it in the first place.

The first thought that comes to my mind is how were these data points collected?  Why are the coordinates for the known location different from the actual coordinates?  Is it due to the accuracy of the GPS device?  Is it a hand-held device as opposed to a survey-level device?  That would explain the discrepancy.

If the discrepancy between the coordinates you obtained and the actual coordinates of the known location are due to the accuracy of the GPS used, you cannot assume that the discrepancy between the coordinates for the other points and the "true" coordinates would be the same in terms of direction and distance.  Every point taken would have its' own level of accuracy in direction and distance.  In fact, dependiing on how many points are involved, the odds are that some of them may be spot on.  Trying to adjust all of the points based on the known location may actually make them less accurate.

That's my $0.02 worth.

Phil
0 Kudos
KenLucas
Occasional Contributor
Phil,
  I appreciate your thoughts on this and agree with your point about the risk of globally shifting all of the points; some could likely move further away from their true location. In this case, the 4 of our points that were taken over survery monuments show us that the error, of the 4,  is consistently 10' northeast of the 4 matching survey monuments. This could be interpreted as the error happended ue to a single surveying mistake.  Where, we don't know. We could simply shift all of the points 10' to the northeast, using the edit selection tool. However, I'd like to take advantage of the apparent consistentcy(less statistical variability) in this case, to shift all of the points mathematically or systematically, using least squares or some other averaging option. I'm hoping the gis community will comment on some of the ways we could shift these points and what other ESRI applications or geoprocessing tools their preferred method would require.  I believe that survey analyst could handle this challenge. However, for those of us who can't afford to purchase Survey A are there other,"poor mans", options for globablly shifting points.
Ken
0 Kudos
TrishRice
Occasional Contributor III
Since you want them all shifted the same it sounds like the easiest way is to do it is Excel like you suggested.

The only way I can think to do it in Map, you'd want the true location of your monuments as a spatial file.  Select all your field data points at once.  With Snapping turned on click and drag one point that corresponds to a control monument over to its true location.  All selected points will move.  This works great for linear movement but not rotation.
0 Kudos
T__WayneWhitley
Frequent Contributor
I'm thinking you have access to the Spatial Adjustment tools, correct?  ...there's a toolbar for that - I haven't used it in awhile and I do not have 9.3.  So I think you need to search the relevant webhelp -- as a matter of fact I stumbled upon this this demonstration:

http://webhelp.esri.com/arcgisdesktop/9.3/tutorials/editor/Editor_Ex8_1.htm

Pretty cool... and I've done so much of this type thing in the past I sometimes forget the difference between some of these tools.  Basically I think this holds true:

- The tools on the 'georeferencing' toolbar are good for transforming image tiles and CAD.
- The tools on the 'spatial adjustment' toolbar are good for feature classes and shapefiles, etc., essentially ESRI's feature or 'vector' data.

Hope that helps... I'm sure there is a good link in the 9.3 webhelp as well.

Enjoy,
Wayne

EDIT:  Ahhh, I remember this - here it is, the relevant 9.3 webhelp ref on spatial adjustment, and it is actually quite good info:

An overview of spatial adjustment
Release 9.3
Last modified May 14, 2009
Editing and data compilation > Spatial adjustment
http://webhelp.esri.com/arcgisdesktop/9.3/index.cfm?id=703&pid=702&topicname=An_overview_of_spatial_...
0 Kudos
TrishRice
Occasional Contributor III
- The tools on the 'georeferencing' toolbar are good for transforming image tiles and CAD.
- The tools on the 'spatial adjustment' toolbar are good for feature classes and shapefiles, etc., essentially ESRI's feature or 'vector' data.


O_o  Aha!  I thought it was odd that georeferencing doesn't work on vectors.  New toy!  Thanks!
0 Kudos
KenLucas
Occasional Contributor
I tried the feature to point tool and couldn't get it to work, ie place a point in the center of the multipoint features.  I input 3 points in a feature class, "points",  and left the box in the tool unchecked. The tool created 3 points, but they are all just coincident with the original points; it did not set a point in the center of the 3 points, as I expected and need it to.   I hope someone can tell me what I'm doing wrong. I also tried it with the box checked.
     However, encouraged by the Wayne and Tish's comments, I did some trial and error and would like to throw these procedures out to the community to comment on. I'd still, really, like to find out the way to use ArcGis to make a least squares or similar weighted average of a set of point.

Procedure to adjust a traverse to remove error indicated by a tie in with a survey monument. Finding and changing a bad bearing or distance.
ArcInfo license/COGO toolbar and list of traverse bearings (deg min seconds or dec degrees format) & distances required. Create a line feature [the cogo tool req a line feature to create a traverse] �?? start editing the line feature �?? editor options �?? set to north azimuth & dec degrees �?? activate the COGO traverse tool - set the start point �?? enter all of the traverse shots, ending at the tie to the survey monument �?? right click on a traverse entry, in the traverse window, that you believe has an inaccurate bearing or distance �?? properties �?? change the bearing or distance [ this will adjust the whole traverse] �?? evaluate how much closer the end of the traverse is to the tied survey monument �?? repeat with other traverse shots until you�??ve eliminated as much error as possible and the end of the traverse is as close to the survey monument as poss. When it�??s close enough to the survery monument finish the traverse. 
Procedure to adjust a series of side shots to a corrected traverse.   These are side shots made off of the original traverse, as adjusted above. Set a point at the end of the adjusted traverse, the point nearest the survey monument �?? create a feature class of all of the side shot points you think need to shift �?? start editing the feature class �?? set snapping to vertex - Select all of the side shot points �?? use the editor tool to drag the side shot points onto the traverse point
Ken
N
0 Kudos
T__WayneWhitley
Frequent Contributor
Ken, I'm not saying this is the way to go to ultimately solve your problem, but if you want to try another shot at Feature to Point at least to have a look at it, you must 1st have true multipoint input to feed into it, i.e. a point 'group' per feature record.  How to get this?  Use of the Dissolve tool is one way, using a common ID by which to group.  Or if you don't have that ID coded right away and just want see what I mean, then select in a map a small cluster of points to experiment with (they can be randomly selected for purposes of this experiment), open Dissolve and enter the point feature layer you have a selection on and run Dissolve (without specifying a field to dissolve by)....check the output- open the attribute table, and it should be a single record.  Selecting the record will highlight the whole group (the points are now 'parts' of a different kind of 'whole', called multipoint).  Input that into Feature to Point and it should give you the centroid of the multipoint geometry.

Hope that's clear.


Enjoy,
Wayne
0 Kudos