Opened 13 years ago
Closed 12 years ago
#1615 closed defect (fixed)
Regression - v.transform fails to create a 3D map, crashes on Windows
| Reported by: | marisn | Owned by: | |
|---|---|---|---|
| Priority: | major | Milestone: | 6.4.3 |
| Component: | Vector | Version: | 6.4.2 |
| Keywords: | Cc: | ||
| CPU: | Unspecified | Platform: | Unspecified |
Description
When running v.transform with zshift option to create a 3D map, there are three issues:
- GUI layout has changed since 6.4.0 (this is no-no for a stable release)
- when executed from WXGUI, there's a simple crash
- when executed from tcltk GUI, it is hanging till dbf.exe process is killed
- resulting map is a 2D instead of 3D, as in 6.4.0
Command used to test:
v.transform input=Dores_punkti_10k@bulba output=dores_punkti_3d table=Dores_punkti_10k columns=zshift:h
Crash info:
Problem Event Name: APPCRASH Application Name: v.transform.exe Application Version: 0.0.0.0 Application Timestamp: 4f410e6e Fault Module Name: ntdll.dll Fault Module Version: 6.0.6002.18541 Fault Module Timestamp: 4ec3e3d5 Exception Code: c0000005 Exception Offset: 00067daa OS Version: 6.0.6002.2.2.0.256.6 Locale ID: 1062 Additional Information 1: 8ed9 Additional Information 2: 802ed84e5e7e3c072dd89c95c6b2c33f Additional Information 3: d541 Additional Information 4: 78eb81a857c38b14810170133ea43f63
Marking as major, as existing functionality was broken in a stable release.
2d/3d issue is most likely introduced by Martin in r45103
Attachments (4)
Change History (27)
comment:1 by , 12 years ago
follow-up: 3 comment:2 by , 12 years ago
It doesn't crash any more, as there have been some DBF related fixes, still produced output is wrong -> it doesn't shift on Z axis, as it was doing in 6.4.0.
v.random output=rm_a_map n=10 zmax=100.0 column='h double precision' v.transform input=rm_a_map output=rm_a_map3d table=rm_a_map columns=zshift:h v.info map=rm_a_map3d
follow-up: 6 comment:3 by , 12 years ago
Replying to marisn:
It doesn't crash any more, as there have been some DBF related fixes, still produced output is wrong -> it doesn't shift on Z axis, as it was doing in 6.4.0.
v.random output=rm_a_map n=10 zmax=100.0 column='h double precision' v.transform input=rm_a_map output=rm_a_map3d table=rm_a_map columns=zshift:h v.info map=rm_a_map3d
tested here with the latest osgeo4w-wingrass6.4.3svn on a win7-64bit-box in the nc-sample dataset:
v.random output=rm_a_map2 n=10 zmax=100.0 column=h v.transform --verbose input=rm_a_map@user1 output=rm_a_map3d table=rm_a_map column="zshift:h"
v.transform crashes, but rm_a_map3d - although corrupted - is created and 3d
corrputed rm_a_map3d
v.build --verbose map=rm_a_map3d@user1 Coor-Datei der Vektorkarte <rm_a_map3d@user1> ist größer als sie sein sollte (304 Byte zuviel).
but 3d
v.info Layer: rm_a_map3d@user1 [...] Karte ist 3D: Yes
Helmut
follow-up: 5 comment:4 by , 12 years ago
I was testing with: WinGRASS-6.4.3svn-r52715-323 No crash and no 3D output. Egggoggg! Now I retested and still no crash, but I see "coor" mismatch indicating failure. Most likely issue is still present, still "crash" part sometimes works? Strange.
On Linux, "coor" is just fine, still map isn't 3D.
comment:5 by , 12 years ago
Replying to marisn:
I was testing with: WinGRASS-6.4.3svn-r52715-323
testing osgeo4w-wingrass6.4.3svn is r52715.
now tested with standalone wingrass6.4.3svn-r52715
No crash and no 3D output.
no crash and no 3D output also here...
v.transform --verbose input=rm_a_map2@user1 output=rm_a_map3d table=rm_a_map column="zshift:h" Erstelle Topologie für die Vektorkarte <rm_a_map3d>... Registriere Primitive... 10 Primitive registriert 10 Vertices registriert Erzeuge Flächen... 0 Flächen angelegt 0 Inseln angelegt Füge Inseln hinzu... Füge Zentroide hinzu... Die Topologie wurde erstellt. Anzahl der Knoten: 10 Anzahl der Primitive: 10 Anzahl der Punkte: 10 Anzahl der Linien: 0 Anzahl der Grenzen: 0 Anzahl der Zentroide: 0 Anzahl der Flächen: 0 Anzahl der Inseln: 0 Neue Vektorkarte <rm_a_map3d> Grenzkoordinaten: N: 227210.852 S: 215367.504 O: 644605.853 W: 632004.608 B: 0.000 T: 0.000 v.transform komplett.
Egggoggg! Now I retested and still no crash, but I see "coor" mismatch indicating failure. Most likely issue is still present, still "crash" part sometimes works? Strange.
??
Helmut
follow-up: 7 comment:6 by , 12 years ago
follow-up: 8 comment:7 by , 12 years ago
follow-up: 9 comment:8 by , 12 years ago
Replying to mmetz:
Fixed in all branches (r52782-4). Please test.
There are more fixes in r52791-3. These should fix the previously observed crashes and the coor mismatch, it was the by now usual suspect of a still open database driver.
For trunk, I have removed that rather weird table option. The table is now selected by layer number or name, as is common for other modules. The example provided in this ticket also uses the attribute table attached to the test vector, but the user has to figure out first the table name, and then supply that table name as argument. The test command for trunk is thus now
v.transform input=rm_a_map output=rm_a_map3d layer=1 column="zshift:h"
Markus M
follow-ups: 10 13 comment:9 by , 12 years ago
Replying to mmetz:
Replying to mmetz:
Fixed in all branches (r52782-4). Please test.
There are more fixes in r52791-3. These should fix the previously observed crashes and the coor mismatch, it was the by now usual suspect of a still open database driver.
tested here with osgeo4w-wingrass6.4.3svn r52917
v.transform still crashes here; the coor mismatch remains, but the output vector seems to be 3D.
v.info --verbose map=rm_a_map3d@user1 D1/3: Vect_open_old(): name = rm_a_map3d@user1 mapset= update = 0 D1/3: Vect_set_thresh(): thresh = 0.000000 D3/3: dig_init_plus() D1/3: dig_spidx_init() D3/3: dig_cidx_init() D1/3: open format file: 'user1/vector/rm_a_map3d/frmt' D1/3: Vector format: 0 (native) D1/3: Vect_set_thresh(): thresh = 0.000000 D1/3: Vect__read_head(): vector = rm_a_map3d@user1 D1/3: Vect_set_thresh(): thresh = 0.000000 D1/3: Level request = 0 D1/3: Vect_open_topo(): name = rm_a_map3d mapset= user1 D1/3: Topo file for vector 'rm_a_map3d@user1' not available. D1/3: Vect_open_old(): vector opened on level 1 D1/3: Vect_read_dblinks(): map = rm_a_map3d, mapset = user1 D3/3: Searching for FID column in OGR DB D1/3: dbln file: \grassdata/nc_spm_08/user1/vector/rm_a_map3d/dbln D1/3: dbln: 1 rm_a_map3d cat $GISDBASE/$LOCATION_NAME/$MAPSET/dbf/ dbf D3/3: Field number <1>, name <(null)> D3/3: Vect_check_dblink: field 1 D1/3: field = 1 name = (null), table = rm_a_map3d, key = cat, database = $GISDBASE/$LOCATION_NAME/$MAPSET/dbf/, driver = dbf D1/3: Dblinks read D1/3: Vect_close(): name = rm_a_map3d, mapset = user1, format = 0, level = 1 D1/3: close history file D1/3: Vect_open_old(): name = rm_a_map3d@user1 mapset= update = 0 D1/3: Vect_set_thresh(): thresh = 0.000000 D3/3: dig_init_plus() D1/3: dig_spidx_init() D3/3: dig_cidx_init() D1/3: open format file: 'user1/vector/rm_a_map3d/frmt' D1/3: Vector format: 0 (native) D1/3: Vect_set_thresh(): thresh = 0.000000 D1/3: Vect__read_head(): vector = rm_a_map3d@user1 D1/3: Vect_set_thresh(): thresh = 0.000000 D1/3: Level request = 1 D1/3: V1_open_old_nat(): name = rm_a_map3d mapset = user1 D2/3: Coor header: file version 5.1 , supported from GRASS version 5.1 D2/3: byte order 0 D2/3: header size 14 D2/3: with_z 224 D2/3: coor size 0 D1/3: get coor info: \grassdata/nc_spm_08/user1/vector/rm_a_map3d/coor D1/3: Info->size = 304, Info->mtime = 1345981666 D1/3: coor size in head = 0, real coor file size= 304 Coor-Datei der Vektorkarte <rm_a_map3d@user1> ist größer als sie sein sollte (304 Byte zuviel). +----------------------------------------------------------------------------+ | Layer: rm_a_map3d@user1 | | Mapset: user1 | | Location: nc_spm_08 | | Datenbank: \grassdata | | Titel: | | Maßstab: 1:1 | | Kartenformat: native | | Name des Erzeugers: myricaria | | Organisation: | | Quelldatum: Sun Aug 26 13:47:46 2012 | |----------------------------------------------------------------------------| | Typ der Karte: Vektor (Level: 1) | | | | Anzahl Punkte: 1 Anzahl Flächen: 0 | | Anzahl Linien: 0 Anzahl Inseln: 0 | | Anzhal Grenzen: 0 Anzahl faces: 0 | | Anzahl Zentroid: 0 Anzahl Kerne: 0 | | | | Karte ist 3D: Yes | | Anzahl der dblinks: 1 | | | | Projektion: Lambert Conformal Conic | | N: 221154.04217658 S: 221154.04217658 | | E: 635205.84734642 W: 635205.84734642 | | B: 0 T: 0 | | | | Digitalisierungs-Schwellwert: 0 | | Kommentare: | | | +----------------------------------------------------------------------------+ (Sun Aug 26 13:49:31 2012) Befehl ausgeführt (0 sec)
v.transform-DEBUG-3-log will be attached.
Helmut
follow-up: 11 comment:10 by , 12 years ago
Replying to hellik:
Replying to mmetz:
Replying to mmetz:
Fixed in all branches (r52782-4). Please test.
There are more fixes in r52791-3. These should fix the previously observed crashes and the coor mismatch, it was the by now usual suspect of a still open database driver.
tested here with osgeo4w-wingrass6.4.3svn r52917
v.transform still crashes here; the coor mismatch remains, but the output vector seems to be 3D.
| Anzahl Punkte: 1 Anzahl Flächen: 0 | Karte ist 3D: Yes | | | | Projektion: Lambert Conformal Conic | | N: 221154.04217658 S: 221154.04217658 | | E: 635205.84734642 W: 635205.84734642 | | B: 0 T: 0 |
Unfortunately, the z coordinate of the one single point created is zero, I would call the output vector fake 3D.
Tested here on XP with the wingrass 6.4.svn standalone installer r52917, should be the same like yours, but it works. No crash, no coor mismatch, all 10 points have been transformed, and bottom and top make sense, i.e. are identical to min and max in the h column.
Markus M
follow-up: 12 comment:11 by , 12 years ago
Replying to mmetz:
Tested here on XP with the wingrass 6.4.svn standalone installer r52917, should be the same like yours, but it works. No crash, no coor mismatch, all 10 points have been transformed, and bottom and top make sense, i.e. are identical to min and max in the h column.
Markus M
now tested with the standalone WinGRASS-6.4.3svn-r52917-329-Setup.exe; no crash and no coor mismatch also here, but it seems no 3D output.
v.transform --verbose input=rm_a_map@user1 output=rm_a_map3d table=rm_a_map column="zshift:h" Erstelle Topologie für die Vektorkarte <rm_a_map3d>... Registriere Primitive... 10 Primitive registriert 10 Vertices registriert Erzeuge Flächen... 0 Flächen angelegt 0 Inseln angelegt Füge Inseln hinzu... Füge Zentroide hinzu... Die Topologie wurde erstellt. Anzahl der Knoten: 10 Anzahl der Primitive: 10 Anzahl der Punkte: 10 Anzahl der Linien: 0 Anzahl der Grenzen: 0 Anzahl der Zentroide: 0 Anzahl der Flächen: 0 Anzahl der Inseln: 0 Neue Vektorkarte <rm_a_map3d> Grenzkoordinaten: N: 228243.324 S: 216493.088 O: 643187.201 W: 630043.489 B: 0.000 T: 0.000 v.transform komplett.
v.transform-input:
v.report map=rm_a_map@user1 option=coor Zeige Typ/Name der Spalten der Datenbank des Layers 1: cat|h|x|y|z 1|64.41847|631050.141911069|219569.490646077|0 2|21.555223|641187.170018616|228243.324076052|0 3|60.863063|637342.295602283|220380.306405835|0 4|38.547929|640841.090121159|216493.087557604|0 5|60.655538|632119.96826075|217402.783288064|0 6|37.482833|632219.306009094|227814.844203009|0 7|48.179571|633124.790185247|220734.214300974|0 8|90.920743|643187.200537126|218178.579058199|0 9|54.017762|630043.488876003|220686.010315256|0 10|98.339793|642816.400646992|222201.757866146|0
v.transform-output:
v.report map=rm_a_map3d@user1 option=coor Zeige Typ/Name der Spalten der Datenbank des Layers 1: cat|h|x|y|z 1|64.41847|631050.141911069|219569.490646077|0 2|21.555223|641187.170018616|228243.324076052|0 3|60.863063|637342.295602283|220380.306405835|0 4|38.547929|640841.090121159|216493.087557604|0 5|60.655538|632119.96826075|217402.783288064|0 6|37.482833|632219.306009094|227814.844203009|0 7|48.179571|633124.790185247|220734.214300974|0 8|90.920743|643187.200537126|218178.579058199|0 9|54.017762|630043.488876003|220686.010315256|0 10|98.339793|642816.400646992|222201.757866146|0
all my tests in this ticket on a win7-64bit-box.
any idea why there are differences between osgeo4w-wingrass and standalone-wingrass on the same machine and obviously between different windows versions?
Helmut
comment:12 by , 12 years ago
Replying to hellik:
Replying to mmetz:
Tested here on XP with the wingrass 6.4.svn standalone installer r52917, should be the same like yours, but it works. No crash, no coor mismatch, all 10 points have been transformed, and bottom and top make sense, i.e. are identical to min and max in the h column.
Markus M
now tested with the standalone WinGRASS-6.4.3svn-r52917-329-Setup.exe; no crash and no coor mismatch also here, but it seems no 3D output.
Weird.
any idea why there are differences between osgeo4w-wingrass and standalone-wingrass on the same machine and obviously between different windows versions?
Unfortunately no. I tried to test on a Windows 7 64 bit box, but I can not start GRASS because
"Der Prozedureinsprungpunkt "sqlite3_open_v2" wurde in der DLL "sqlite3.dll" nicht gefunden."
Now that is even weirder because GRASS does not use sqlite3_open_v2, instead it uses sqlite3_open which apparently works, but sqlite3_open calls sqlite3_open_v2 which does not work.
Please do not respond to this comment within this ticket because this sqlite issue should be a separate ticket. I am investigating.
Markus M
follow-up: 14 comment:13 by , 12 years ago
Replying to hellik:
Replying to mmetz:
Replying to mmetz:
Fixed in all branches (r52782-4). Please test.
There are more fixes in r52791-3. These should fix the previously observed crashes and the coor mismatch, it was the by now usual suspect of a still open database driver.
tested here with osgeo4w-wingrass6.4.3svn r52917
v.transform still crashes here; the coor mismatch remains, but the output vector seems to be 3D.
tested now several times in osgeo4w-wingrass6.4.3svn r52917 (heavy other work are in the background: arcgis, gimp, etc.): crash or no crash seems to be random; if no crash, no 3D output.
still memory issues?
follow-ups: 15 16 comment:14 by , 12 years ago
Replying to hellik:
tested now several times in osgeo4w-wingrass6.4.3svn r52917 (heavy other work are in the background: arcgis, gimp, etc.): crash or no crash seems to be random; if no crash, no 3D output.
still memory issues?
Could be. In your debug report, only two out of ten attribute table entries are copied, then the debug report ends. But then this is such a small vector, just a few KB, that memory issues are hard to imagine.
Some testing suggestions:
restart Windows, start GRASS and nothing else, then test
v.random output=rm_a_map n=10 zmax=100.0 column="h double precision" v.transform input=rm_a_map output=rm_a_map3d table=rm_a_map columns=zshift:h v.info map=rm_a_map3d
now restart Windows, first start all sorts of other apps, make them work long and hard, now start GRASS with the commands above.
Ideally, do everything with DEBUG=4 and report the results.
No further ideas from my side so far (GRASS is still not starting on my Windows 7 box).
Thanks a lot for your efforts!
Markus M
comment:15 by , 12 years ago
Replying to mmetz:
Some testing suggestions:
restart Windows, start GRASS and nothing else, then test
v.random output=rm_a_map n=10 zmax=100.0 column="h double precision" v.transform input=rm_a_map output=rm_a_map3d table=rm_a_map columns=zshift:h v.info map=rm_a_map3d
testing part 1: restarting windows, notepad++ (for storing debug info) and GRASS as only two started software (norton firewall in the background).
v.random --verbose output=rm_a_map n=10 zmax=100.0 column=h double precision v.report map=rm_a_map@user1 option=coor Zeige Typ/Name der Spalten der Datenbank des Layers 1: cat|h|x|y|z 1|4.629658|635570.696127201|220671.590319529|0 2|83.669546|636487.166966765|216334.467604602|0 3|83.394879|640837.427900021|222324.945829646|0 4|32.062746|635209.967345195|225262.09295938|0 5|59.938353|633464.003418073|217668.935209204|0 6|0.018311|632945.341349529|215628.299813837|0 7|0.012207|632348.399304178|221032.91421247|0 8|56.599628|638098.544267098|222962.309640797|0 9|7.74865|639610.126041444|226608.096560564|0 10|74.4438|634732.505264443|219473.49467452|0
v.transform input=rm_a_map output=rm_a_map3d table=rm_a_map columns=zshift:h
v.transform crashes, debug=4-log as txt-file attached.
v.info-debug=4-log also in the txt-file.
+----------------------------------------------------------------------------+ | Layer: rm_a_map3d@user1 | | Mapset: user1 | | Location: nc_spm_08 | | Datenbank: \grassdata | | Titel: | | Maßstab: 1:1 | | Kartenformat: native | | Name des Erzeugers: myricaria | | Organisation: | | Quelldatum: Sun Aug 26 18:34:00 2012 | |----------------------------------------------------------------------------| | Typ der Karte: Vektor (Level: 1) | | | | Anzahl Punkte: 10 Anzahl Flächen: 0 | | Anzahl Linien: 0 Anzahl Inseln: 0 | | Anzhal Grenzen: 0 Anzahl faces: 0 | | Anzahl Zentroid: 0 Anzahl Kerne: 0 | | | | Karte ist 3D: Yes | | Anzahl der dblinks: 1 | | | | Projektion: Lambert Conformal Conic | | N: 226608.09656056 S: 215628.29981384 | | E: 640837.42790002 W: 632348.39930418 | | B: 0.012207 T: 83.669546 | | | | Digitalisierungs-Schwellwert: 0 | | Kommentare: | | | +----------------------------------------------------------------------------+
v.report map=rm_a_map3d@user1 option=coor ERROR: Kann die Vektorkarte <rm_a_map3d@user1> nicht auf Level 2 öffnen. Versuchen Sie die Topologie mit v.build neu anzulegen. cat|x|y|z
v.build map=rm_a_map3d@user1
v.report map=rm_a_map3d@user1 option=coor Coor-Datei der Vektorkarte <rm_a_map3d@user1> ist größer als sie sein sollte (384 Byte zuviel). cat|x|y|z 1|635570.696127201|220671.590319529|4.629658 2|636487.166966765|216334.467604602|83.669546 3|640837.427900021|222324.945829646|83.394879 4|635209.967345195|225262.09295938|32.062746 5|633464.003418073|217668.935209204|59.938353 6|632945.341349529|215628.299813837|0.018311 7|632348.399304178|221032.91421247|0.012207 8|638098.544267098|222962.309640797|56.599628 9|639610.126041444|226608.096560564|7.74865 10|634732.505264443|219473.49467452|74.4438 (Sun Aug 26 18:38:52 2012) Befehl ausgeführt (1 sec)
so it seems the v.transform is now 3D, but still coor mismatch.
now restart Windows, first start all sorts of other apps, make them work long and hard, now start GRASS with the commands above.
...after next windows restart.
by , 12 years ago
| Attachment: | wingrass643svn_commands_wo_background.txt added |
|---|
debug=4 without other sw in background
by , 12 years ago
| Attachment: | wingrass643svn_commands_w_background.txt added |
|---|
debug=4 with other sw in background
follow-up: 17 comment:16 by , 12 years ago
Replying to mmetz:
now restart Windows, first start all sorts of other apps, make them work long and hard, now start GRASS with the commands above.
restart done, a lot of actions done (gdalwarp,gdaltranslate, ArcGIS-reprojecting, GIMP, R, etc) and all this software still open.
v.report map=rm_a_map@user1 option=coor Zeige Typ/Name der Spalten der Datenbank des Layers 1: cat|h|x|y|z 1|23.679312|642811.822870571|221936.841944639|0 2|35.941649|639653.157139805|219384.09070101|0 3|97.265542|642408.978545488|218689.046906949|0 4|80.935087|637440.260017701|223433.225501267|0 5|99.530015|644367.351298563|217720.023194067|0 6|25.189978|633553.27005829|223358.241523484|0 7|47.856075|641798.760948515|226636.936552019|0 8|61.809137|630641.80425428|228375.164036988|0 9|9.561449|635350.04730369|215969.43571276|0 10|27.835932|643034.302804651|216352.595599231|0
v.transform input=rm_a_map output=rm_a_map3d table=rm_a_map columns=zshift:h
v.transform crashes, debug=4-log as txt-file attached.
v.info-debug=4-log also in the txt-file.
v.info map=rm_a_map3d@user1 +----------------------------------------------------------------------------+ | Layer: rm_a_map3d@user1 | | Mapset: user1 | | Location: nc_spm_08 | | Datenbank: \grassdata | | Titel: | | Maßstab: 1:1 | | Kartenformat: native | | Name des Erzeugers: myricaria | | Organisation: | | Quelldatum: Sun Aug 26 20:46:44 2012 | |----------------------------------------------------------------------------| | Typ der Karte: Vektor (Level: 1) | | | | Anzahl Punkte: 10 Anzahl Flächen: 0 | | Anzahl Linien: 0 Anzahl Inseln: 0 | | Anzhal Grenzen: 0 Anzahl faces: 0 | | Anzahl Zentroid: 0 Anzahl Kerne: 0 | | | | Karte ist 3D: Yes | | Anzahl der dblinks: 1 | | | | Projektion: Lambert Conformal Conic | | N: 228375.16403699 S: 215969.43571276 | | E: 644367.35129856 W: 630641.80425428 | | B: 9.561449 T: 99.530015 | | | | Digitalisierungs-Schwellwert: 0 | | Kommentare: | | | +----------------------------------------------------------------------------+
v.report map=rm_a_map3d@user1 option=coor ERROR: Kann die Vektorkarte <rm_a_map3d@user1> nicht auf Level 2 öffnen. Versuchen Sie die Topologie mit v.build neu anzulegen. cat|x|y|z
v.build map=rm_a_map3d@user1
v.report map=rm_a_map3d@user1 option=coor cat|x|y|z 1|642811.822870571|221936.841944639|23.679312 2|639653.157139805|219384.09070101|35.941649 3|642408.978545488|218689.046906949|97.265542 4|637440.260017701|223433.225501267|80.935087 5|644367.351298563|217720.023194067|99.530015 6|633553.27005829|223358.241523484|25.189978 7|641798.760948515|226636.936552019|47.856075 8|630641.80425428|228375.164036988|61.809137 9|635350.04730369|215969.43571276|9.561449 10|643034.302804651|216352.595599231|27.835932
?
Helmut
follow-ups: 18 20 comment:17 by , 12 years ago
Replying to hellik:
?
could it be some quoting in the command?
A
v.transform input=rm_a_map@user1 output=rm_a_map3d table=rm_a_map column="zshift:h"
vs.
B
v.transform input=rm_a_map output=rm_a_map3d table=rm_a_map columns=zshift:h
A => crashing, coor mismatch, no 3D output B => crashing, coor mismatch, 3D output
?
Helmut
follow-up: 19 comment:18 by , 12 years ago
comment:19 by , 12 years ago
Replying to hellik:
all commands (by me) were given in the wxgui-command-line
cmd.exe- and dbf.exe-zombies in the windows task manager
follow-up: 21 comment:20 by , 12 years ago
Replying to hellik:
Replying to hellik:
?
could it be some quoting in the command?
A
v.transform input=rm_a_map@user1 output=rm_a_map3d table=rm_a_map column="zshift:h"vs.
B
v.transform input=rm_a_map output=rm_a_map3d table=rm_a_map columns=zshift:hA => crashing, coor mismatch, no 3D output B => crashing, coor mismatch, 3D output
tested today with osgeo4w-wingrass6.4.3svn-r52928.
very strange, today no crash and no coor mismatch, but differences if quotes are used in the columns-input (tested in the wxgui-commandline:
v.transform input=rm_a_map output=rm_a_map3di table=rm_a_map columns=zshift:h Erstelle Topologie für die Vektorkarte <rm_a_map3di>... Registriere Primitive... 10 Primitive registriert 10 Vertices registriert Erzeuge Flächen... 0 Flächen angelegt 0 Inseln angelegt Füge Inseln hinzu... Füge Zentroide hinzu... Anzahl der Knoten: 10 Anzahl der Primitive: 10 Anzahl der Punkte: 10 Anzahl der Linien: 0 Anzahl der Grenzen: 0 Anzahl der Zentroide: 0 Anzahl der Flächen: 0 Anzahl der Inseln: 0 Neue Vektorkarte <rm_a_map3di> Grenzkoordinaten: N: 227107.440 S: 215934.828 O: 642462.539 W: 630227.515 B: 0.119 T: 98.663 v.transform komplett.
=> correct 3D output, but
v.transform input=rm_a_map output=rm_a_map3dk table=rm_a_map columns="zshift:h" Erstelle Topologie für die Vektorkarte <rm_a_map3dk>... Registriere Primitive... 10 Primitive registriert 10 Vertices registriert Erzeuge Flächen... 0 Flächen angelegt 0 Inseln angelegt Füge Inseln hinzu... Füge Zentroide hinzu... Anzahl der Knoten: 10 Anzahl der Primitive: 10 Anzahl der Punkte: 10 Anzahl der Linien: 0 Anzahl der Grenzen: 0 Anzahl der Zentroide: 0 Anzahl der Flächen: 0 Anzahl der Inseln: 0 Neue Vektorkarte <rm_a_map3dk> Grenzkoordinaten: N: 227107.440 S: 215934.828 O: 642462.539 W: 630227.515 B: 0.000 T: 0.000 v.transform komplett.
no 3D output.
from the manual (http://grass.osgeo.org/grass64/manuals/html64_user/v.transform.html)
v.transform -t myarchsites output=myarchsites3d column="zshift:zs" table=myarchsites
?
Helmut ?
Helmut
follow-up: 22 comment:21 by , 12 years ago
Replying to hellik:
Replying to hellik:
Replying to hellik:
?
could it be some quoting in the command?
A
v.transform input=rm_a_map@user1 output=rm_a_map3d table=rm_a_map column="zshift:h"vs.
B
v.transform input=rm_a_map output=rm_a_map3d table=rm_a_map columns=zshift:hA => crashing, coor mismatch, no 3D output B => crashing, coor mismatch, 3D output
tested today with osgeo4w-wingrass6.4.3svn-r52928.
very strange, today no crash and no coor mismatch, but differences if quotes are used in the columns-input (tested in the wxgui-commandline:
Looks like the 'columns' option is not recognized. Unrecognized columns answers are silently ignored by v.transform. Warning added in trunk r52937.
from the manual (http://grass.osgeo.org/grass64/manuals/html64_user/v.transform.html)
v.transform -t myarchsites output=myarchsites3d column="zshift:zs" table=myarchsites
AFAIK, quotes are only needed if the answer contains whitespace, and then also only on the command line, not the GUI. Update the manual? Anyway, this seems to be a different issue.
Markus M
follow-up: 23 comment:22 by , 12 years ago
Replying to mmetz:
AFAIK, quotes are only needed if the answer contains whitespace, and then also only on the command line, not the GUI. Update the manual?
maybe an update of the manual, because columns="zshift:h" doesn't work neither in the wxgui-command line nor in the v.transform-gui.
Helmut
p.s. tested today osgeo4w-wingrass6.4.3svn-r52935 several times, no crash, no coormismatch, correct 3D output
comment:23 by , 12 years ago
| Resolution: | → fixed |
|---|---|
| Status: | new → closed |
Please provide the "Dores_punkti_10k" dataset or an equivalent sample.