XYZ File Not Importing

[Anonymous]

All-
I am new to your forum and hoping somebody can help me out. I have provided the attached xyz file of a survey I did for one of my Archicad clients. They are trying to use the new Import Surveyors Data tool in ArchiCad 14 but it is not working. Can somebody try this on their machine and if you can make it work will you please tell me how you did it? and also can you please send me the mesh? The XYZ file is in international feet.

thanks
Thomas Gnauck

[Anonymous]

Where is the attached xyz file ??

[Erika Epstein]

Where is Work?

[Anonymous]

Here is the file iI changed the extension to PDF to pass the filter the file is actually a txt file if you change the extension- The work is in Montana USA

thanks
t.

[Greg A_ Blunier]

I think you need to clean up the text file.
It wouldn't import for me, so I looked at it in a text editor and found:
1. It had 4251 points, many of which were duplicates. After sorting and deleting the duplicates, there were only 758 points.
2. The coordinate values were huge. They were over 300 miles from the origin. I would recommend "offsetting" the values before importing so that they are closer to the origin.
Greg

[Anonymous]

Sorry gnauck,

All my tryes with your file FAILED.....AC14 STOP's RESPONDING. Even if I use only 20 points.
Try to make Mesh from myone attached file.

Send your file to Graphisoft.

[Erika Epstein]

gnauk,
The file you posted is problematic. When I compared it with the a civil txt file I have that works, there is a fourth column filled with
Soft 1
Soft 4
Soft 3

etc.

HTH

[Anonymous]

andro55 wrote:
Sorry gnauck,

All my tryes with your file FAILED.....AC14 STOP's RESPONDING. Even if I use only 20 points.
Try to make Mesh from myone attached file.

Send your file to Graphisoft.

Howdy all,

I'm on the AC end working with gnauck on this. Thanks for all the input thus far.

andro55, I had no problems converting the file you uploaded. Oddly enough, I WAS able to convert approx. 1/4 of gnauck's file (by manually deleting the latter 3/4's of it in the .txt file before converting in AC14) but I just tried to do it again, and even that didn't work.(!!??) With that said, I sent the file to Karl O. to see if a faster processor would do the trick. He couldn't convert it either. Noting that the coordinates are large numbers, Karl took another approach (quoted below):

"AC 14 basically froze up on me with your data file. I had a feeling that the coordinate values were too big for what AC could deal with - it cannot cope with coordinates that are really far from the origin. So, I did a replace all to get rid of the "171" prefix on every line, making whatever that value is smaller. It then converted instantly. I don't know if doing that alters things in a bad way - you'll have to check."

It's looking like this might be a "too far from origin" problem, though I'm also curious as to gnauck's response to Greg's point #1 regarding the duplicate points.

Will be sending this to GS tech support to get their take on it....

[Greg A_ Blunier]

After cleaning out the duplicate points, I was able to get it to import. If you want me to send the .pln, send me a private message with contact details.
Greg

[Karl Ottenstein]

Greg wrote:
After cleaning out the duplicate points, I was able to get it to import. If you want me to send the .pln, send me a private message with contact details.
Greg

Found the same thing, Greg.

Here's how I eliminated the duplicates on the Mac:
Open a Terminal window into the folder where the xyz file resides.
Enter the command:
cat nobelxyz.txt | sort | uniq > nobelxyz_clean.txt

then use the nobelxyz_clean.txt file.

It seems fishy that reducing the x coord value also allowed generation of a mesh, even with duplicates. And, the freeze-thing shows that GS needs to check the validity of the input file better and gracefully produce a message. So, this should still get sent to tech support.

Attached is the no-duplicates file.

Cheers,
Karl