javaseis-devel Mailing List for JavaSeis
Status: Pre-Alpha
Brought to you by:
chuckmosher
Menu
▾
▴
javaseis-devel — Developers mailing list
You can subscribe to this list here.
| 2005 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
(2) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
(14) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2006 |
Jan
(1) |
Feb
(1) |
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2010 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
(2) |
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Bill M. <bil...@gm...> - 2010-06-14 19:35:13
|
Chuck, Do you have any plans for creating a gcj build? Just thinking about how nice it would be to link in from some C++ code to an already-built library. Bill M. Chuck Mosher wrote: > Greetings all ! > > I'll be undertaking a number of projects related to JavaSeis this summer, so want to alert you to upcoming changes. I plan to keep a running commentary on what I'm up to on the SourceForge JavaSeis developer's forum: > > https://sourceforge.net/projects/javaseis/forums/forum/397964 > > I just completed a check-in of some changes and refactoring to the org.javaseis.array and org.javaseis.parallel packages, so take a look and let me know what you think. > > Also, since the mailing list and forum has been inactive for 4 years (?!), let me know if you're no longer interested in participating and I will remove you from the project and mailing lists. > > -Chuck > > ------------------------------------------------------------------------------ > ThinkGeek and WIRED's GeekDad team up for the Ultimate > GeekDad Father's Day Giveaway. ONE MASSIVE PRIZE to the > lucky parental unit. See the prize list and enter to win: > http://p.sf.net/sfu/thinkgeek-promo > _______________________________________________ > Javaseis-devel mailing list > Jav...@li... > https://lists.sourceforge.net/lists/listinfo/javaseis-devel > |
|
From: Chuck M. <chu...@ya...> - 2010-06-14 16:11:16
|
Greetings all ! I'll be undertaking a number of projects related to JavaSeis this summer, so want to alert you to upcoming changes. I plan to keep a running commentary on what I'm up to on the SourceForge JavaSeis developer's forum: https://sourceforge.net/projects/javaseis/forums/forum/397964 I just completed a check-in of some changes and refactoring to the org.javaseis.array and org.javaseis.parallel packages, so take a look and let me know what you think. Also, since the mailing list and forum has been inactive for 4 years (?!), let me know if you're no longer interested in participating and I will remove you from the project and mailing lists. -Chuck |
|
From: Chuck M. <chu...@ya...> - 2006-02-11 20:18:25
|
I'm pleased to report that Dave Diller has repaired the SMART interfaces, and I have completed the DistributedArray and SeisFft3d classes. The snapshot referred to above (Snapshot_021106) now provides a valid "build" that can be used for testing these classes.
In particular, I would appreciate any feedback on the DistributedArray classes.
The code still has a way to go, however. The next set of major changes will include:
- Updates from Bill Lucas once he completes a set of SeisSpace read/write modules
- Refactoring the Seisio classes to split out TraceMaps that can be used for sorting
- Adding serial and parallel sort classes
- Load and Store methods for moving DistributedArrays to/from disk
- Service interfaces for coordinates, binning grids, and velocities
But it's nice to see some of the major elements settling down.
Let me know if you have opinions on the design of the next set of classes, or would like to take them on.
-Chuck
|
|
From: Steve A. <san...@lg...> - 2006-01-24 17:39:32
|
All,
Am I overlooking something or are there problems with the latest
snapshot? I'm getting errors about incompatible types and symbols that
can't be found.
Thanks,
Steve
compile:
[javac] Compiling 52 source files to /data/devtools/javaseis/classes
[javac] /data/devtools/javaseis/src/jsTestPanel.java:457:
incompatible types
[javac] found : long[]
[javac] required: int[]
[javac] int[] axisLengths =
jsFile.getGridDefinition().getAxisLengths();
[javac]
^
[javac] /data/devtools/javaseis/src/jsTestPanel.java:675: cannot
find symbol
[javac] symbol : method getHeaderDefinition()
[javac] location: class org.javaseis.io.Seisio
[javac] TraceProperties headerDefinition =
jsFile.getHeaderDefinition();
[javac] ^
[javac] /data/devtools/javaseis/src/jsTestPanel.java:782: cannot
find symbol
[javac] symbol : method getHeaderDefinition()
[javac] location: class org.javaseis.io.Seisio
[javac] TraceProperties headerDefinition =
jsFile.getHeaderDefinition();
[javac] ^
[javac] /data/devtools/javaseis/src/jsTestPanel.java:895: cannot
find symbol
[javac] symbol : variable RECEIVER
[javac] location: class org.javaseis.properties.AxisLabel
[javac] axisLabels[2] = AxisLabel.RECEIVER;
[javac] ^
[javac]
/data/devtools/javaseis/src/org/javaseis/smart/JavaSeisDataContext.java:263:
cannot find symbol
[javac] symbol : method getAxisUnitsStrings()
[javac] location: class org.javaseis.smart.SmartGridDefinition
[javac] return _smartGridDefinition.getAxisUnitsStrings();
[javac] ^
[javac]
/data/devtools/javaseis/src/org/javaseis/smart/JavaSeisDataContext.java:273:
cannot find symbol
[javac] symbol : method getAxisDomainsStrings()
[javac] location: class org.javaseis.smart.SmartGridDefinition
[javac] return _smartGridDefinition.getAxisDomainsStrings();
[javac] ^
[javac]
/data/devtools/javaseis/src/org/javaseis/smart/JavaSeisDataContext.java:288:
cannot find symbol
[javac] symbol : method getAxisLogicalOrigins()
[javac] location: class org.javaseis.smart.SmartGridDefinition
[javac] return _smartGridDefinition.getAxisLogicalOrigins();
[javac] ^
[javac]
/data/devtools/javaseis/src/org/javaseis/smart/JavaSeisDataContext.java:293:
cannot find symbol
[javac] symbol : method getAxisLogicalDeltas()
[javac] location: class org.javaseis.smart.SmartGridDefinition
[javac] return _smartGridDefinition.getAxisLogicalDeltas();
[javac] ^
[javac]
/data/devtools/javaseis/src/org/javaseis/smart/JavaSeisDataContext.java:298:
cannot find symbol
[javac] symbol : method getAxisPhysicalOrigins()
[javac] location: class org.javaseis.smart.SmartGridDefinition
[javac] return _smartGridDefinition.getAxisPhysicalOrigins();
[javac] ^
[javac]
/data/devtools/javaseis/src/org/javaseis/smart/JavaSeisDataContext.java:303:
cannot find symbol
[javac] symbol : method getAxisPhysicalDeltas()
[javac] location: class org.javaseis.smart.SmartGridDefinition
[javac] return _smartGridDefinition.getAxisPhysicalDeltas();
[javac] ^
[javac]
/data/devtools/javaseis/src/org/javaseis/smart/SmartFile.java:55: cannot
find symbol
[javac] symbol : constructor
Seisio(java.lang.String,org.javaseis.smart.SmartGridDefinition,org.javaseis.properties.DataDefinition,org.javaseis.properties.TraceProperties)
[javac] location: class org.javaseis.io.Seisio
[javac] super(path,
[javac] ^
[javac]
/data/devtools/javaseis/src/org/javaseis/smart/SmartFile.java:83: cannot
find symbol
[javac] symbol : constructor
Seisio(java.lang.String,org.javaseis.smart.SmartGridDefinition,org.javaseis.properties.DataDefinition,org.javaseis.properties.TraceProperties)
[javac] location: class org.javaseis.io.Seisio
[javac] super(path,
[javac] ^
[javac]
/data/devtools/javaseis/src/org/javaseis/smart/SmartFile.java:106:
cannot find symbol
[javac] symbol : method getSmartGridDefinition()
[javac] location: class org.javaseis.smart.SmartFile
[javac] SmartGridDefinition smartGridDefinition =
getSmartGridDefinition();
[javac] ^
[javac]
/data/devtools/javaseis/src/org/javaseis/smart/SmartFile.java:109:
cannot find symbol
[javac] symbol : variable gridDefinition
[javac] location: class org.javaseis.smart.SmartFile
[javac] dataDefinition, gridDefinition, headerDefinition,
opaqueDataNames, opaqueDataObjects);
[javac] ^
[javac]
/data/devtools/javaseis/src/org/javaseis/smart/SmartFileTests.java:57:
cannot find symbol
[javac] symbol : constructor
SmartGridDefinition(int,org.javaseis.properties.AxisDefinition[])
[javac] location: class org.javaseis.smart.SmartGridDefinition
[javac] SmartGridDefinition smartGridDefinition = new
SmartGridDefinition(numDimensions, axes);
[javac] ^
[javac]
/data/devtools/javaseis/src/org/javaseis/smart/SmartFileTests.java:107:
cannot find symbol
[javac] symbol : method getSmartGridDefinition()
[javac] location: class org.javaseis.smart.SmartFile
[javac] int nval =
(int)(svf1.getSmartGridDefinition().getNumSamplesPerTrace()
[javac] ^
[javac]
/data/devtools/javaseis/src/org/javaseis/smart/SmartFileTests.java:108:
cannot find symbol
[javac] symbol : method getSmartGridDefinition()
[javac] location: class org.javaseis.smart.SmartFile
[javac] *
svf1.getSmartGridDefinition().getNumTracesPerFrame());
[javac] ^
[javac] Note: Some input files use unchecked or unsafe operations.
[javac] Note: Recompile with -Xlint:unchecked for details.
[javac] 17 errors
----------------------------------------------------------------------
This e-mail, including any attached files, may contain confidential and privileged information for the sole use of the intended recipient. Any review, use, distribution, or disclosure by others is strictly prohibited. If you are not the intended recipient (or authorized to receive information for the intended recipient), please contact the sender by reply e-mail and delete all copies of this message.
|
|
From: Chuck M. <chu...@ya...> - 2005-12-28 21:23:30
|
I've added the first of several draft interfaces that are designed to support JavaSeis Parallel Distributed Objects. The first is a simple message passing interface and an implementation using MPI Java. The new package is "org.javaseis.parallel". The next interface will define regularly sampled JavaSeis parallel arrays. That should be enough to produce an example 3D FFT using JavaSeis parallel arrays. -Chuck |
|
From: Chuck M. <chu...@ya...> - 2005-12-23 22:14:07
|
One more checkin this week. Refactored the Seisio class to use the SeisProperties class rather than duplicate code in Seisio. This also required changes to SmartFile and SmartFileTests. -Chuck |
|
From: Chuck M. <chu...@ya...> - 2005-12-22 15:31:13
|
After Dave's changes, I did some modest refactoring of the Seisio and Virtualio classes to add static delete methods, cleaned up the Virtualio constructors to remove unused arguments, and moved the AxisLabel class from "smart" to "properties" (to remove references to SMART from Seisio). All tests run at this point, and the Seisio tests now remove the directories used for the test. -Chuck |
|
From: Done, B. <Bil...@co...> - 2005-12-20 15:53:51
|
Tag it anyway. Tags are cheap. The tag name must start with a letter and may then contain letters, numbers, the hyphen, and the underscore. Bill Bill Done Subsurface Technology Work Phone: 281.293.4589 ConocoPhillips Work FAX : 281.293.2015 600 N. Dairy Ashford Cell Phone: 918.261.3561 Houston, TX 77079-1175 Work email: Bil...@co... -----Original Message----- From: jav...@li... [mailto:jav...@li...] On Behalf Of Chuck Mosher Sent: Tuesday, December 20, 2005 7:58 AM To: jav...@li... Cc: Lucas, Bill A Subject: [Javaseis-devel] Tests after Dave's checkin Tests are running now after Dave's latest checkin (output below). Next task on my list is to begin adding support for JavaSeis style "Parallel Distributed Objects". Might even come up with a better name. This will likely require some refactoring of the "smart" interfaces, so we might want to branch or tag at this point and provide a "snapshot" of the current code state. Bill Lucas - let me know if you want to check in any changes before we tag a current snapshot. -Chuck |
|
From: Chuck M. <chu...@ya...> - 2005-12-20 13:58:07
|
Tests are running now after Dave's latest checkin (output below). Next task on my list is to begin adding support for JavaSeis style "Parallel Distributed Objects". Might even come up with a better name. This will likely require some refactoring of the "smart" interfaces, so we might want to branch or tag at this point and provide a "snapshot" of the current code state. Bill Lucas - let me know if you want to check in any changes before we tag a current snapshot. -Chuck ---------------------------------------------------------------------------- org.javaseis.smart.SmartFileTests ***** SUCCESS ***** SeisCompress SCALAR16 validation completed SeisCompress NATIVE validation completed Seisio: Could not delete path /home/moshecc/svfTestDataset/TraceProperties Seisio: Created 'raw' file /home/moshecc/svfTestDataset Seisio: Completed 'raw' file write Seisio: Opened 'raw' file for read Seisio: Completed 'raw' file validation Seisio: Created 'virtual' file directories Seisio: Created 'mapped virtual' file /home/moshecc/svfTestDataset Seisio: reading from disk at offset 0 for volume index 0 Seisio: writing to disk at offset 0 for volume index 0 Seisio: writing to disk at offset 0 for volume index 0 Seisio: reading from disk at offset 84 for volume index 1 Seisio: writing to disk at offset 84 for volume index 1 Seisio: writing to disk at offset 84 for volume index 1 Seisio: reading from disk at offset 168 for volume index 2 Seisio: writing to disk at offset 168 for volume index 2 Seisio: writing to disk at offset 168 for volume index 2 Seisio: reading from disk at offset 252 for volume index 3 Seisio: writing to disk at offset 252 for volume index 3 Seisio: writing to disk at offset 252 for volume index 3 Seisio: reading from disk at offset 336 for volume index 4 Seisio: writing to disk at offset 336 for volume index 4 Seisio: writing to disk at offset 336 for volume index 4 Seisio: reading from disk at offset 420 for volume index 5 Seisio: writing to disk at offset 420 for volume index 5 Seisio: writing to disk at offset 420 for volume index 5 Seisio: reading from disk at offset 504 for volume index 6 Seisio: writing to disk at offset 504 for volume index 6 Seisio: writing to disk at offset 504 for volume index 6 Seisio: reading from disk at offset 588 for volume index 7 Seisio: writing to disk at offset 588 for volume index 7 Seisio: writing to disk at offset 588 for volume index 7 Seisio: reading from disk at offset 672 for volume index 8 Seisio: writing to disk at offset 672 for volume index 8 Seisio: writing to disk at offset 672 for volume index 8 Seisio: reading from disk at offset 756 for volume index 9 Seisio: writing to disk at offset 756 for volume index 9 Seisio: writing to disk at offset 756 for volume index 9 Seisio: reading from disk at offset 840 for volume index 10 Seisio: Completed 'mapped virtual' file write Seisio: writing to disk at offset 840 for volume index 10 Seisio: Opened 'mapped virtual' file for read Seisio: reading from disk at offset 0 for volume index 0 Seisio: reading from disk at offset 84 for volume index 1 Seisio: reading from disk at offset 168 for volume index 2 Seisio: reading from disk at offset 252 for volume index 3 Seisio: reading from disk at offset 336 for volume index 4 Seisio: reading from disk at offset 420 for volume index 5 Seisio: reading from disk at offset 504 for volume index 6 Seisio: reading from disk at offset 588 for volume index 7 Seisio: reading from disk at offset 672 for volume index 8 Seisio: reading from disk at offset 756 for volume index 9 Seisio: reading from disk at offset 840 for volume index 10 Seisio: Completed 'mapped virtual' file validation #www.javaseis.org - JavaSeis TraceProperites 2005.0 #Tue Dec 20 07:42:05 CST 2005 SourceZ="Source Z coord,DOUBLE,1,24" SourceY="Source Y coord,DOUBLE,1,16" InLine="In-Line number,INT,1,0" SourceX="Source X coord,DOUBLE,1,8" X-Line="X-Line number,INT,1,4" ReceiverZ="Receiver Z coord,DOUBLE,1,48" ReceiverY="Receiver Y coord,DOUBLE,1,40" ReceiverX="Receiver X coord,DOUBLE,1,32" InLine,X-Line,Source XYZ, Receiver XYZ Index 0 InLine 200 X-Line 100 Source X 0.0 Receiver X 0.0 Index 1 InLine 200 X-Line 100 Source X 50.0 Receiver X -50.0 Index 2 InLine 200 X-Line 100 Source X 100.0 Receiver X -100.0 Index 3 InLine 200 X-Line 100 Source X 150.0 Receiver X -150.0 Index 4 InLine 200 X-Line 100 Source X 200.0 Receiver X -200.0 Index 5 InLine 200 X-Line 100 Source X 250.0 Receiver X -250.0 Index 6 InLine 200 X-Line 100 Source X 300.0 Receiver X -300.0 Index 7 InLine 200 X-Line 100 Source X 350.0 Receiver X -350.0 Index 8 InLine 200 X-Line 100 Source X 400.0 Receiver X -400.0 Index 9 InLine 200 X-Line 100 Source X 450.0 Receiver X -450.0 Dec 20, 2005 7:42:05 AM org.javaseis.smart.SmartFrameworkTest testSanity INFO: SmartFrameworkTest.testSanity ***** SUCCESS ***** org.javaseis.smart.SmartFrameworkTest ***** SUCCESS ***** |
|
From: Dave D. <DD...@lg...> - 2005-12-19 21:31:34
|
Bill,
=20
I think that I am finished with the org.javaseis code, for at least a
week or two.
=20
I fixed org.javaseis.smart.SmartFrameworkTest so that it runs. I had to
disable
the trace header part of the tests - support for trace headers has
either been
removed or morphed into something that I don't recognize.
=20
Dave
________________________________
From: Dave Diller=20
Sent: Sunday, December 18, 2005 6:03 PM
To: jav...@li...
Cc: 'Mosher, Chuck C.'; 'Bil...@co...'; Dan Grygier;
Amelia Webster
Subject: I merged in my JavaSeis changes
=20
Folks,
=20
I merged in my changes to the JavaSeis code. Most of them are related
to making the data
type and the axis labels enums instead of static Strings. I won't let
that happen again, my
list of top 5 rules includes "Don't Work On Branched Code".
=20
Bill Lucas - my merge was rather painful, and yours will probably be
worse. Please let me
know if you want some help.
=20
NOTE: The autotests did not all run when I fetched a fresh copy of the
code on
Saturday, and they still don't all run (I didn't fix any of them - shame
on me). See
details below.
=20
Dave
=20
The following classes have main() methods that run tests:
=20
java org.javaseis.compress.SeisCompress
java org.javaseis.io.Seisio
java org.javaseis.properties.SeisTraceProperties
java org.javaseis.smart.SmartFileTests
java -ea org.javaseis.smart.SmartFrameworkTest
=20
Three of them fail, as follows:
=20
java org.javaseis.compress.SeisCompress
SeisCompress SCALAR16 validation completed
Exception in thread "main" java.nio.BufferOverflowException
at
java.nio.DirectFloatBufferS.put(DirectFloatBufferS.java:311)
at
org.javaseis.compress.SeisCompress.packTrace(SeisCompress.java:180)
at
org.javaseis.compress.SeisCompress.packFrame(SeisCompress.java:168)
at
org.javaseis.compress.SeisCompress.main(SeisCompress.java:363)
=20
java org.javaseis.io.Seisio
Seisio: Could not delete path /home/dave/svfTestDataset/FileProperties
Exception in thread "main" org.javaseis.io.SeisioException: VFIO_DIR0 =3D
/home/dave/tmp/0 is not a directory
at org.javaseis.io.Virtualio.openVirtual(Virtualio.java:274)
at org.javaseis.io.Virtualio.<init>(Virtualio.java:163)
at org.javaseis.io.Seisio.delete(Seisio.java:827)
at org.javaseis.io.Seisio.main(Seisio.java:1687)
=20
java org.javaseis.smart.SmartFileTests
SmartFileTests: Created 'raw' file SmartDataset1
Exception in thread "main" java.lang.RuntimeException:
Seisio.getTraceHeaderArray() is not implemented
at
org.javaseis.io.Seisio.getTraceHeaderArray(Seisio.java:1190)
at
org.javaseis.smart.SmartFile.getAllTraceHeaders(SmartFile.java:331)
at
org.javaseis.smart.SmartFileTests.chucksOriginalTest(SmartFileTests.java
:117)
at
org.javaseis.smart.SmartFileTests.main(SmartFileTests.java:319)
=20
----------------------------------------------------------------------
This e-mail, including any attached files, may contain confidential and pri=
vileged information for the sole use of the intended recipient. Any review=
, use, distribution, or disclosure by others is strictly prohibited. If yo=
u are not the intended recipient (or authorized to receive information for =
the intended recipient), please contact the sender by reply e-mail and dele=
te all copies of this message. |
|
From: Dave D. <DD...@lg...> - 2005-12-19 18:20:56
|
I just create a class org.javaseis.smart.AxisLabel, because Strings were a bit too ambiguous and an enum was too limiting. As a side benefit, the axis label has a name and a description. I think that this coupling of the name and description may address the confusion mentioned below. The way I am using it, the AxisLabel name corresponds to a header name and the description is a user-familiar string. -----Original Message----- From: jav...@li... [mailto:jav...@li...] On Behalf Of Chuck Mosher Sent: Sunday, December 18, 2005 8:38 AM To: jav...@li... Subject: Re: [Javaseis-devel] Offset as a SMART Axis ? Dave, =20 We need to make a distinction between header names and axis names. "Offset" is the name of the axis, "SEQNO" or "OFB_NO" are examples of ProMAX headers that might be used as the "unique" header for that axis. =20 This has been a bit confusing for users. We have a three-level header renumbering program that is used quite often to create a set of integer headers that uniquely number the Volume:Frame:Trace axes. I experimented with a set of headers that were explicitly called "VOLUME", "FRAME", and "TRACE", but after a transpose they don't make sense, of course. =20 We have a pretty good sense of the headers that are used data in Sail Line or Shot order: S_LINE, SOURCE, CHAN =20 For CDP data our current set of standard headers is: ILINE_NO, XLINE_NO, SEQNO =20 But this changes to: OFB_NO, ILINE_NO, XLINE_NO =20 for common offset data. Since data are accessed as frames, the index within the frame is not critical, so the use of SEQNO has not been a big deal. =20 This also brings up the option of including other header types for indexing, such as floats or longs. It's been easier to just deal with ints so far. =20 By the way, just finished a billion trace sort this weekend. Required 128 nodes to get it to work, due to 32 bit indexing limits. An int runs out of room pretty quick when you have a billion traces. This is a case that illustrates why it might be useful to have longs for indexing. =20 -Chuck ----- Original Message ---- From: Chuck Mosher <chu...@ya...> To: jav...@li... Sent: Sunday, December 18, 2005 9:35:10 AM Subject: [Javaseis-devel] Offset as a SMART Axis ? Chuck, I don't see how source-receiver offset can be a SMART axis, since it is not an integer value. Seems like it would have to be offset bin. Dave ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_idv37&alloc_id865&op=3Dclick _______________________________________________ Javaseis-devel mailing list Jav...@li... https://lists.sourceforge.net/lists/listinfo/javaseis-devel ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_idv37&alloc_id=16865&op=3Dick _______________________________________________ Javaseis-devel mailing list Jav...@li... https://lists.sourceforge.net/lists/listinfo/javaseis-devel ---------------------------------------------------------------------- This e-mail, including any attached files, may contain confidential and pri= vileged information for the sole use of the intended recipient. Any review= , use, distribution, or disclosure by others is strictly prohibited. If yo= u are not the intended recipient (or authorized to receive information for = the intended recipient), please contact the sender by reply e-mail and dele= te all copies of this message. |
|
From: Dave D. <DD...@lg...> - 2005-12-19 18:11:08
|
I agree with Chuck, but my view is slanted by the need for compatibility with SeisSpace/ProMAX. I might add that the "packed" set of arbitrary object has been a powerful feature over the years. All of the code in org.javaseis.smart uses longs (64-bit integers) for indices. I would not want to use a double, because it always comes down to integer arithmetic when dealing with offsets in arrays and files. -----Original Message----- From: jav...@li... [mailto:jav...@li...] On Behalf Of Chuck Mosher Sent: Monday, December 19, 2005 7:52 AM To: jav...@li... Subject: [Javaseis-devel] JavaSeis Trace Properties Bill, =20 With the NIO package, we can extend the concept of a trace header from an array of a particular type to a "packed" set of arbitrary objects. We are using the terminology "TraceProperties" as the analog for trace headers. I think the flexibility of arbitrary data types has advantages over a fixed type. The main disadvantage is that when defining and using a TraceProperty, care must be given to consider future uses, such as using a 32 bit int for a trace index. =20 If we do a good job on refactoring the current JavaSeis reference implementation to define a set of interfaces, then it should be possible to support a variety of "Trace Header" formats, including SEGY and the heritage Conoco 64 bit model. =20 For now we will stick to the current implementation that is more aligned with the ProMAX model, but will include ideas like you have presented as design points for refactoring to interfaces. =20 -Chuck ----- Original Message ---- From: Bill Menger <bil...@gm...> Cc: jav...@li... Sent: Sunday, December 18, 2005 9:55:39 PM Subject: Re: [Javaseis-devel] I merged in my JavaSeis changes Regarding Chuck's email concerning headers, I would like consideration=20 given to making all headers 64 bit doubles. Conoco has used these (or=20 their Cray 60-bit equivalents) since 1986 and hasn't had header word=20 overflow issues yet. I realize that integer arithmetic can be quicker=20 but don't think this has ever been an issue. ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log files for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_id=3D7637&alloc_id=3D16865&op=3Dclick _______________________________________________ Javaseis-devel mailing list Jav...@li... https://lists.sourceforge.net/lists/listinfo/javaseis-devel ---------------------------------------------------------------------- This e-mail, including any attached files, may contain confidential and pri= vileged information for the sole use of the intended recipient. Any review= , use, distribution, or disclosure by others is strictly prohibited. If yo= u are not the intended recipient (or authorized to receive information for = the intended recipient), please contact the sender by reply e-mail and dele= te all copies of this message. |
|
From: Chuck M. <chu...@ya...> - 2005-12-19 14:52:38
|
Bill, With the NIO package, we can extend the concept of a trace header from an array of a particular type to a "packed" set of arbitrary objects. We are using the terminology "TraceProperties" as the analog for trace headers. I think the flexibility of arbitrary data types has advantages over a fixed type. The main disadvantage is that when defining and using a TraceProperty, care must be given to consider future uses, such as using a 32 bit int for a trace index. If we do a good job on refactoring the current JavaSeis reference implementation to define a set of interfaces, then it should be possible to support a variety of "Trace Header" formats, including SEGY and the heritage Conoco 64 bit model. For now we will stick to the current implementation that is more aligned with the ProMAX model, but will include ideas like you have presented as design points for refactoring to interfaces. -Chuck ----- Original Message ---- From: Bill Menger <bil...@gm...> Cc: jav...@li... Sent: Sunday, December 18, 2005 9:55:39 PM Subject: Re: [Javaseis-devel] I merged in my JavaSeis changes Regarding Chuck's email concerning headers, I would like consideration given to making all headers 64 bit doubles. Conoco has used these (or their Cray 60-bit equivalents) since 1986 and hasn't had header word overflow issues yet. I realize that integer arithmetic can be quicker but don't think this has ever been an issue. |
|
From: Bill M. <bil...@gm...> - 2005-12-19 03:55:36
|
Regarding Chuck's email concerning headers, I would like consideration given to making all headers 64 bit doubles. Conoco has used these (or their Cray 60-bit equivalents) since 1986 and hasn't had header word overflow issues yet. I realize that integer arithmetic can be quicker but don't think this has ever been an issue. Dave Diller wrote: > Folks, > > I merged in my changes to the JavaSeis code. Most of them are related > to making the data > > type and the axis labels enums instead of static Strings. I won’t let > that happen again, my > > list of top 5 rules includes “Don’t Work On Branched Code”. > > Bill Lucas – my merge was rather painful, and yours will probably be > worse. Please let me > > know if you want some help. > > NOTE: The autotests did not all run when I fetched a fresh copy of the > code on > > Saturday, and they still don’t all run (I didn’t fix any of them – > shame on me). See > > details below. > > Dave > > The following classes have main() methods that run tests: > > java org.javaseis.compress.SeisCompress > > java org.javaseis.io.Seisio > > java org.javaseis.properties.SeisTraceProperties > > java org.javaseis.smart.SmartFileTests > > java -ea org.javaseis.smart.SmartFrameworkTest > > Three of them fail, as follows: > > java org.javaseis.compress.SeisCompress > > SeisCompress SCALAR16 validation completed > > Exception in thread "main" java.nio.BufferOverflowException > > at java.nio.DirectFloatBufferS.put(DirectFloatBufferS.java:311) > > at org.javaseis.compress.SeisCompress.packTrace(SeisCompress.java:180) > > at org.javaseis.compress.SeisCompress.packFrame(SeisCompress.java:168) > > at org.javaseis.compress.SeisCompress.main(SeisCompress.java:363) > > java org.javaseis.io.Seisio > > Seisio: Could not delete path /home/dave/svfTestDataset/FileProperties > > Exception in thread "main" org.javaseis.io.SeisioException: VFIO_DIR0 > = /home/dave/tmp/0 is not a directory > > at org.javaseis.io.Virtualio.openVirtual(Virtualio.java:274) > > at org.javaseis.io.Virtualio.<init>(Virtualio.java:163) > > at org.javaseis.io.Seisio.delete(Seisio.java:827) > > at org.javaseis.io.Seisio.main(Seisio.java:1687) > > java org.javaseis.smart.SmartFileTests > > SmartFileTests: Created 'raw' file SmartDataset1 > > Exception in thread "main" java.lang.RuntimeException: > Seisio.getTraceHeaderArray() is not implemented > > at org.javaseis.io.Seisio.getTraceHeaderArray(Seisio.java:1190) > > at org.javaseis.smart.SmartFile.getAllTraceHeaders(SmartFile.java:331) > > at > org.javaseis.smart.SmartFileTests.chucksOriginalTest(SmartFileTests.java:117) > > at org.javaseis.smart.SmartFileTests.main(SmartFileTests.java:319) > > ------------------------------------------------------------------------ > This e-mail, including any attached files, may contain confidential > and privileged information for the sole use of the intended recipient. > Any review, use, distribution, or disclosure by others is strictly > prohibited. If you are not the intended recipient (or authorized to > receive information for the intended recipient), please contact the > sender by reply e-mail and delete all copies of this message. -- No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.1.371 / Virus Database: 267.14.1/206 - Release Date: 12/16/2005 |
|
From: Dave D. <DD...@lg...> - 2005-12-19 01:03:40
|
Folks,
=20
I merged in my changes to the JavaSeis code. Most of them are related
to making the data
type and the axis labels enums instead of static Strings. I won't let
that happen again, my
list of top 5 rules includes "Don't Work On Branched Code".
=20
Bill Lucas - my merge was rather painful, and yours will probably be
worse. Please let me
know if you want some help.
=20
NOTE: The autotests did not all run when I fetched a fresh copy of the
code on
Saturday, and they still don't all run (I didn't fix any of them - shame
on me). See
details below.
=20
Dave
=20
The following classes have main() methods that run tests:
=20
java org.javaseis.compress.SeisCompress
java org.javaseis.io.Seisio
java org.javaseis.properties.SeisTraceProperties
java org.javaseis.smart.SmartFileTests
java -ea org.javaseis.smart.SmartFrameworkTest
=20
Three of them fail, as follows:
=20
java org.javaseis.compress.SeisCompress
SeisCompress SCALAR16 validation completed
Exception in thread "main" java.nio.BufferOverflowException
at
java.nio.DirectFloatBufferS.put(DirectFloatBufferS.java:311)
at
org.javaseis.compress.SeisCompress.packTrace(SeisCompress.java:180)
at
org.javaseis.compress.SeisCompress.packFrame(SeisCompress.java:168)
at
org.javaseis.compress.SeisCompress.main(SeisCompress.java:363)
=20
java org.javaseis.io.Seisio
Seisio: Could not delete path /home/dave/svfTestDataset/FileProperties
Exception in thread "main" org.javaseis.io.SeisioException: VFIO_DIR0 =3D
/home/dave/tmp/0 is not a directory
at org.javaseis.io.Virtualio.openVirtual(Virtualio.java:274)
at org.javaseis.io.Virtualio.<init>(Virtualio.java:163)
at org.javaseis.io.Seisio.delete(Seisio.java:827)
at org.javaseis.io.Seisio.main(Seisio.java:1687)
=20
java org.javaseis.smart.SmartFileTests
SmartFileTests: Created 'raw' file SmartDataset1
Exception in thread "main" java.lang.RuntimeException:
Seisio.getTraceHeaderArray() is not implemented
at
org.javaseis.io.Seisio.getTraceHeaderArray(Seisio.java:1190)
at
org.javaseis.smart.SmartFile.getAllTraceHeaders(SmartFile.java:331)
at
org.javaseis.smart.SmartFileTests.chucksOriginalTest(SmartFileTests.java
:117)
at
org.javaseis.smart.SmartFileTests.main(SmartFileTests.java:319)
=20
----------------------------------------------------------------------
This e-mail, including any attached files, may contain confidential and pri=
vileged information for the sole use of the intended recipient. Any review=
, use, distribution, or disclosure by others is strictly prohibited. If yo=
u are not the intended recipient (or authorized to receive information for =
the intended recipient), please contact the sender by reply e-mail and dele=
te all copies of this message. |
|
From: Chuck M. <chu...@ya...> - 2005-12-18 15:38:21
|
Dave, =20 We need to make a distinction between header names and axis names. "Offset" is the name of the axis, "SEQNO" or "OFB_NO" are examples of ProMA= X headers that might be used as the "unique" header for that axis. =20 This has been a bit confusing for users. We have a three-level header renum= bering program that is used quite often to create a set of integer headers = that uniquely number the Volume:Frame:Trace axes. I experimented with a set of headers that were explicitly called "VOLUME", = "FRAME", and "TRACE", but after a transpose they don't make sense, of cours= e. =20 We have a pretty good sense of the headers that are used data in Sail Line = or Shot order: S_LINE, SOURCE, CHAN =20 For CDP data our current set of standard headers is: ILINE_NO, XLINE_NO, SEQNO =20 But this changes to: OFB_NO, ILINE_NO, XLINE_NO =20 for common offset data. Since data are accessed as frames, the index within= the frame is not critical, so the use of SEQNO has not been a big deal. =20 This also brings up the option of including other header types for indexing= , such as floats or longs. It's been easier to just deal with ints so far. =20 By the way, just finished a billion trace sort this weekend. Required 128 n= odes to get it to work, due to 32 bit indexing limits. An int runs out of r= oom pretty quick when you have a billion traces. This is a case that illust= rates why it might be useful to have longs for indexing. =20 -Chuck ----- Original Message ---- From: Chuck Mosher <chu...@ya...> To: jav...@li... Sent: Sunday, December 18, 2005 9:35:10 AM Subject: [Javaseis-devel] Offset as a SMART Axis ? Chuck, I don=E2=80=99t see how source-receiver offset can be a SMART axis, since i= t is not an integer value. Seems like it would have to be offset bin. Dave ------------------------------------------------------- This SF.net email is sponsored by: Splunk Inc. Do you grep through log file= s for problems? Stop! Download the new AJAX search engine that makes searching your log files as easy as surfing the web. DOWNLOAD SPLUNK! http://ads.osdn.com/?ad_idv37&alloc_id865&op=3Dclick _______________________________________________ Javaseis-devel mailing list Jav...@li... https://lists.sourceforge.net/lists/listinfo/javaseis-devel |
|
From: Chuck M. <chu...@ya...> - 2005-12-18 15:35:17
|
Chuck, =20 I don=E2=80=99t see how source-receiver offset can be a SMART axis, since i= t is not an integer value. Seems like it would have to be offset bin. =20 Dave |
|
From: Chuck M. <chu...@ya...> - 2005-12-18 15:20:36
|
Welcome to the JavaSeis developers mailing list. Please use this list as much as you can to host discussions about JavaSeis design, features, and other issues. Hopefully, as we populate this discussion we will build up material for FAQs and other docs. Best Regards, Chuck Mosher |
|
From: Chuck M. <chu...@ya...> - 2005-07-24 23:23:18
|
The class names and packages on SourceForge have been updated to reflect the new package heirarchy we discussed last week: org.javaseis.io Virtualio (old namd Seisio) I/O with virtualization across multiple files Seisio (old name Svfio) Seismic dataset management SeisioException(SvfioException) SeisioFileNames(SvfFileNames) org.javaseis.compress SeisCompress Compression 16 bit, 8 bit, gzip, ... org.javaseis.smart Smart*.java GridDefinition SmartDataContext Dave Diller's SMART layer org.javaseis.properties SeisTraceProperty, SeisPropertyHandler(SeisTraceProperties), SeisPropertyType File and trace property definitions and handlers |
|
From: Chuck M. <chu...@ya...> - 2005-07-24 20:46:10
|
Welcome to the JavaSeis Developer's Mailing list, hosted at SourceForge.net. Please use this mailing list for communication regarding development activities in the JavaSeis project. Regards, Chuck Mosher |
