Ticket #1785 (closed defect: worksforme)

Opened 6 years ago

Last modified 6 years ago

1.4.1 Regression while reading MISR data

Reported by: frankie Owned by: warmerdam
Priority: normal Milestone: 1.4.3
Component: GDAL_Raster Version: 1.4.1
Severity: normal Keywords: hdf4
Cc:

Description (last modified by warmerdam) (diff)

See difference among 1.3.2 and 1.4.1 gdalinfo reporting for a MISR (HDF4-EOS sort of) data file: 

#
--- gdal-1.3.2.txt      2007-09-03 16:35:16.000000000 +0200
#
+++ gdal-1.4.1.txt      2007-09-03 16:34:32.000000000 +0200
#
@@ -1,10 +1,10 @@
#
 Driver: HDF4Image/HDF4 Dataset
#
-Size is -1, -1
#
+Size is 0, 0
#
 Coordinate System is:
#
 LOCAL_CS["GCTP projection number 22",
#
     UNIT["Meter",1]]
#
-Origin = (7460750.000000,1090650.000000)
#
-Pixel Size = (1100.00000000,-1100.00000000)
#
+Origin = (7460750.000000000000000,1090650.000000000000000)
#
+Pixel Size = (1100.000000000000000,-1100.000000000000000)
#
 Metadata:
#
   HDFEOSVersion=HDFEOS_V2.13
#
   Path_number=196
#
@@ -367,16 +367,16 @@
#
 
#
 END
#
 
#
-  Block_size.resolution_x=1100, -1221782280, 170812526, 0
#
-  Block_size.resolution_y=1100, -1221782280, 170812526, 0
#
-  Block_size.size_x=128, -1221782280, 170812526, 0
#
-  Block_size.size_y=512, -1221782280, 170812526, 0
#
-  Scale factor=0.0472032241523266, 2.08961414804199e-76, 6.38706223130238e-67, 2.61433546175396e-310, 0, 0, 0, 0
#
-  std_solar_wgted_height=1867.27, -1.031649e-05, 8.396684e-33, 0
#
-  SunDistanceAU=1.01511331019503, 3.33825405993698e+151, 1.66707796419125e-71, 1.05220814586365e-153, 3.53541941098601e-57, 2.37636271514038e-38, 3.27740040683152e+179, 2.61433545205249e-310
#
+  Block_size.resolution_x=1100, -1221287568, 170812526, 0
#
+  Block_size.resolution_y=1100, -1221287568, 170812526, 0
#
+  Block_size.size_x=128, -1221287568, 170812526, 0
#
+  Block_size.size_y=512, -1221287568, 170812526, 0
#
+  Scale factor=0.0472032241523266, 6.52630755860018e-38, 6.38706223130238e-67, 2.61433546175396e-310, 0, 0, 0, 0
#
+  std_solar_wgted_height=1867.27, -1.076642e-05, 8.396684e-33, 0
#
+  SunDistanceAU=1.01511331019503, 3.33825405993698e+151, 1.66707796419125e-71, 1.05220814586365e-153, 3.53854064208632e-57, 2.37636271514037e-38, 3.27740040683152e+179, 2.61433545205249e-310
#
 Corner Coordinates:
#
 Upper Left  ( 7460750.000, 1090650.000)
#
-Lower Left  ( 7460750.000, 1091750.000)
#
-Upper Right ( 7459650.000, 1090650.000)
#
-Lower Right ( 7459650.000, 1091750.000)
#
-Center      ( 7460200.000, 1091200.000)
#
+Lower Left  ( 7460750.000, 1090650.000)
#
+Upper Right ( 7460750.000, 1090650.000)
#
+Lower Right ( 7460750.000, 1090650.000)
#
+Center      ( 7460750.000, 1090650.000)

Specifically vertical dimension collapses and r.in.gdal is not too happy to see that, it warns and creates an invalid location.

WARNING: G_set_window(): North must be larger than South

PS: I can provide data on demand, feel free to ping me about that.

Change History

Changed 6 years ago by frankie

Sorry resubmitting for formatting 

#
--- gdal-1.3.2.txt      2007-09-03 16:35:16.000000000 +0200
#
+++ gdal-1.4.1.txt      2007-09-03 16:34:32.000000000 +0200
#
@@ -1,10 +1,10 @@
#
 Driver: HDF4Image/HDF4 Dataset
#
-Size is -1, -1
#
+Size is 0, 0
#
 Coordinate System is:
#
 LOCAL_CS["GCTP projection number 22",
#
     UNIT["Meter",1]]
#
-Origin = (7460750.000000,1090650.000000)
#
-Pixel Size = (1100.00000000,-1100.00000000)
#
+Origin = (7460750.000000000000000,1090650.000000000000000)
#
+Pixel Size = (1100.000000000000000,-1100.000000000000000)
#
 Metadata:
#
   HDFEOSVersion=HDFEOS_V2.13
#
   Path_number=196
#
@@ -367,16 +367,16 @@
#
 
#
 END
#
 
#
-  Block_size.resolution_x=1100, -1221782280, 170812526, 0
#
-  Block_size.resolution_y=1100, -1221782280, 170812526, 0
#
-  Block_size.size_x=128, -1221782280, 170812526, 0
#
-  Block_size.size_y=512, -1221782280, 170812526, 0
#
-  Scale factor=0.0472032241523266, 2.08961414804199e-76, 6.38706223130238e-67, 2.61433546175396e-310, 0, 0, 0, 0
#
-  std_solar_wgted_height=1867.27, -1.031649e-05, 8.396684e-33, 0
#
-  SunDistanceAU=1.01511331019503, 3.33825405993698e+151, 1.66707796419125e-71, 1.05220814586365e-153, 3.53541941098601e-57, 2.37636271514038e-38, 3.27740040683152e+179, 2.61433545205249e-310
#
+  Block_size.resolution_x=1100, -1221287568, 170812526, 0
#
+  Block_size.resolution_y=1100, -1221287568, 170812526, 0
#
+  Block_size.size_x=128, -1221287568, 170812526, 0
#
+  Block_size.size_y=512, -1221287568, 170812526, 0
#
+  Scale factor=0.0472032241523266, 6.52630755860018e-38, 6.38706223130238e-67, 2.61433546175396e-310, 0, 0, 0, 0
#
+  std_solar_wgted_height=1867.27, -1.076642e-05, 8.396684e-33, 0
#
+  SunDistanceAU=1.01511331019503, 3.33825405993698e+151, 1.66707796419125e-71, 1.05220814586365e-153, 3.53854064208632e-57, 2.37636271514037e-38, 3.27740040683152e+179, 2.61433545205249e-310
#
 Corner Coordinates:
#
 Upper Left  ( 7460750.000, 1090650.000)
#
-Lower Left  ( 7460750.000, 1091750.000)
#
-Upper Right ( 7459650.000, 1090650.000)
#
-Lower Right ( 7459650.000, 1091750.000)
#
-Center      ( 7460200.000, 1091200.000)
#
+Lower Left  ( 7460750.000, 1090650.000)
#
+Upper Right ( 7460750.000, 1090650.000)
#
+Lower Right ( 7460750.000, 1090650.000)
#
+Center      ( 7460750.000, 1090650.000)

Changed 6 years ago by warmerdam

  • keywords hdf4 added
  • version changed from unspecified to 1.4.1
  • component changed from default to GDAL_Raster
  • description modified (diff)
  • milestone set to 1.4.3

Changed 6 years ago by warmerdam

  • status changed from new to assigned

I have tried the following command on Linux with GDAL 1.4 branch (roughly 1.4.2 for these purposes) and got reasonable bounds output. 

 gdalinfo 'HDF4_EOS:EOS_GRID:"MISR_GRP_ELLIPSOID_GM_DF_F03_0024.A2003160.103657.D0105.C006.hdf":BlueBand:Blue Radiance/RDQI'
...
Corner Coordinates:
Upper Left  ( 7460750.000, 1090650.000) 
Lower Left  ( 7460750.000,  527450.000) 
Upper Right ( 7601550.000, 1090650.000) 
Lower Right ( 7601550.000,  527450.000) 
Center      ( 7531150.000,  809050.000)

Could you be more specific about which subdataset you are getting the problematic results?

Changed 6 years ago by frankie

gdalinfo HDF4_EOS:EOS_GRID:"MISR_GRP_ELLIPSOID_GM_DF_F03_0024.A2003160.103657.D0105.C006.hdf":BlueBand:Blue

For your information I got the same result with 1.4.2

Changed 6 years ago by warmerdam

Frankie,

I'm sorry, I'm not clear whether you mean you got the same results that I did or whether you got the same erroneous results. Also, your pasted command does not seem to include the full name "Blue Radiance/RDQI" that was in mine. Is that intentional? Quoting of subdataset names on the commandline can also be very important to prevent the shell from doing funky stuff to names with quotes and spaces.

Changed 6 years ago by warmerdam

  • status changed from assigned to closed
  • resolution set to worksforme

frankie: feel free to close the report, indeed i used the wrong sds tag :-/

Note: See TracTickets for help on using tickets.