-
Notifications
You must be signed in to change notification settings - Fork 2
OlegSmirnov
This page is a running log of my activities. I'm also breaking out specific projects into separate pages, beginning with the 3C147 reductions: OlegSmirnov/3C147. Here is a .TipsAndTricks page I keep for my own benefit.
23/09/2014: OlegSmirnov/GoogleCloudExperiments.
11/04/2014: Got instructions all about Packaging from Gijs.
14/12/2012: BuildingCasaCoreAndAoFlaggerOnUbuntu1204.
04/06/2011: More QMC2 curiosities. Processing the "8-way mispointed" observation of Mar 23, 2011. http://www.astron.nl/meqwiki-data/users/oms/QMC2/QMC2.Mar23-8way.purrlog/
21/03/2001: More investigations of beamshapes using the QMC/QMC2 data: http://www.astron.nl/meqwiki-data/users/oms/QMC2/QMC2.beamshapes.purrlog/
25/02/2011: APERTIF simulations here: http://www.astron.nl/meqwiki-data/users/oms/Apertif-Element-Gain-Drifts-Simulations.purrlog/
24/10/2010: FITS beams now working inside Siamese. http://www.astron.nl/meqwiki-data/users/oms/fits-beams.purrlog/
01/09/2010: CALIM2010 football photos: football_photos.tar
12/08/2010: Ghostbusters! http://www.astron.nl/meqwiki-data/users/oms/Ghostbusters.purrlog
22/07/2010: Playing with KAT-4 data: http://www.astron.nl/meqwiki-data/users/oms/KAT4/KAT4-CenA.purrlog
12/07/2010: Some fascinating results from the QMC and QMC2 field, see purrlogs here:
- QMC2 Jul 21 (pointing errors) observation: http://www.astron.nl/meqwiki-data/users/oms/QMC2/QMC2.2010Jul21.purrlog/
- QMC2 Jul 4 ("error-free") observation: http://www.astron.nl/meqwiki-data/users/oms/QMC2/QMC2.2010Jul4.purrlog/
- QMC observation: http://www.astron.nl/meqwiki-data/users/oms/QMC/qmc.purrlog/index.html. 07/07/2010: Still struggling to get time-averaging of MSs under control. casa 3.0.2 does something, but it is SLOOOOOOW. NDPPP is proving a pain to build, see .BuildingNdppp.
04/03/2010: Releasing an Owlcat tutorial: http://www.astron.nl/meqwiki-data/users/oms/Owlcat-plotms-tutorial.purrlog/
13/02/2010: http://www.astron.nl/meqwiki-data/users/oms/EJones/NVSS-candidate-search.purrlog. Looking for a candidate field for an E-Jones testing observation with the WSRT.
09/10/2009: http://www.astron.nl/meqwiki-data/users/oms/purrlog-SkyJonesVis/
16/06/2009: DependencyActionItems
08/06/2009: PurrInstallationNotes
30/03/2009: UvBrickValidation.
12/03/2009: OxfordWorkshopNotes.
16/12/2008: NewRepositoryLayout, DocumentationListForMarch.
25/11/2008: Need a VellsLabelingScheme for plotters.
07/10/2008: Beginning .
18/09/2008: Renewed CalibrationEfforts
7/07/2008: Added the makems tool, build instructions here: BuildingMakeMs.
2/06/2008: Moving to casacore: LinkingWithCasaCore
21/06/2007: Started a page on StandardContracts.
07/06/2007: Revised handling of TdlOptions.
16/03/2007: Started ResultsLog.
11/02/2007: Started LsmWishList.
Mirroring aips++ for Hans: tar zcvf /home/oms/aips++-weekly.tgz --exclude /weekly/doc/ data/catalogs data/ephemerides data/geodetic data/gui pgplot weekly
02/11/2006: .MimExperiments.
13/09/2006: Started a page on .
03/08/2006: .PlansForSecondHalf2006.
29/04/2006: Some notes on .FringeFitting.
29/04/2006: Sarod requested numarray version 1.5+, which is currently in Debian/testing only. Upgraded this on lofar9 via apt-get install -t testing python2.3-numarray
, this installed fine needing no dependences, so we'll leave it at this version for now and see how it works.
29/04/2006: My ClarSimulations experiences.
17/04/2006: I suspect the old .PythonCrashes were due to a funklet bug (http://lofar9.astron.nl/bugzilla/show_bug.cgi?id=357 and http://lofar9.astron.nl/bugzilla/show_bug.cgi?id=337). I have therefore reinstalled the Debian stable packages for PyQt, sip4, etc. on lofar10 and birch. The only thing still built from source is PyQwt. Will see if we experience any more problems. Here are the currently installed packages:
- {{{# apt-show-versions | grep python2.3 python2.3-sip4-qt3/stable uptodate 4.1.1-1 python2.3-sip4-dev/stable uptodate 4.1.1-1 python2.3-qt3-gl/stable uptodate 3.13-4 python2.3-qt3/stable uptodate 3.13-4 python2.3-qtext/stable uptodate 3.13-4 python2.3-dev/stable upgradeable from 2.3.5-3 to 2.3.5-3sarge1 python2.3/stable upgradeable from 2.3.5-3 to 2.3.5-3sarge1 python2.3-numeric/stable uptodate 23.8-1 python2.3-numarray/stable uptodate 1.1.1-3 python2.3-opengl/stable uptodate 2.0.1.08-5.1 # apt-show-versions | grep libqt libqt3c102-mt/stable uptodate 3:3.3.4-3 libqthreads-12/stable uptodate 1.6.7-1 libqt3-mt-dev/stable uptodate 3:3.3.4-3 libqt3-headers/stable uptodate 3:3.3.4-3 }}}
14/04/2006: Started .PortingTimbaTo64bit.
06/02/2006: Useful article on memory usage in Linux: http://virtualthreads.blogspot.com/2006/02/understanding-memory-usage-on-linux.html
28/01/2006: Started a page on .ClusteringIssues.
04/01/2006: Revisions to NewStreamControl.
15/12/2005: Some notes on .CountedRefsAndThreadSafety.
13/12/2005: still struggling with python crashes. Will rebuild everything from Python and Qt up. See .PythonCrashes.
22/11/2005: libtool-1.5.20 not compatible with LOFAR build. Must install libtool-1.5.6 from stable! See also /etc/apt/preferences
on lofar10 to pin it to thisversion.
15/11/2005: Debian-testing's buggered up Python/PyQt somehow, and meqbrowser.py keeps crashing. Debian-stable, however, is incompatible with PyQwt due to a missing pyqtconfig.py (should be in python2.3-qt3, but isn't... GRRRRR!!!) Current solution as implemented on birch:
- deinstall all python2.3 and install python2.3 python2.3-dev python2.3-numarray from stable ``` apt-get remove python2.3 python-qt-dev apt-get -t stable install python2.3 python2.3-dev python2.3-numarray
* install Qt-3.3.4, etc. from stable ```
apt-get remove qt3-dev-tools libqt3-headers
apt-get -t stable install libqt3-mt-dev libqt3-headers libqscintilla-dev
- here's the current versions on birch: ``` birch# apt-show-versions | grep python python/stable uptodate 2.3.5-2 python2.3-dev/stable uptodate 2.3.5-3 python2.3/stable uptodate 2.3.5-3 birch# python2.3-numarray/stable uptodate 1.1.1-3
libqt3c102-mt/stable uptodate 3:3.3.4-3 libqt3-mt-dev/stable uptodate 3:3.3.4-3 libqt3-headers/stable uptodate 3:3.3.4-3
* Remove sip4 package (`apt-get remove sip4`) and everything related to it
* Build sip4, [[PyQt|PyQt]] and [[PyQwt|PyQwt]] from sources (copy from `birch:/usr/local/src`)
* For sip4: `export QTDIR=/usr/share/qt3 && python configure.py && make && make install`
* For [[PyQt|PyQt]]: `python configure.py -n /usr/include/qscintilla -o /usr/lib && make && make install`
* For [[PyQwt|PyQwt]]: `python configure.py && make && make install`
##### Misc. notes and links
Installing [[../DebianOnTheLatitudeX1|DebianOnTheLatitudeX1]].
[[../DownloadingTools|DownloadingTools]].
##### core.pid files
When pthread_create is invoked by a program that later dumps core, core files become named core.pid automatically. Can't find any way around it. Program below demonstrates this succintly. So, to avoid polluting the home filesystem with core dumps, I do:
* ```
echo /tmp/core.%u >/proc/sys/kernel/core_pattern
which causes all cores to be named /tmp/core.uid
or /tmp/core.uid.pid
.
- {{{/* starts a thread and aborts. core file will be named core.pid! */ #include <pthread.h> #include <stdlib.h>
void * test (void *t) {
- return 0; }
int main ( int argc,const char *argv[] ) {
- pthread_t tid; tid = pthread_create(&tid,0,test,0); abort(); return 0; } }}}
StatCVS stats here: http://lofar9/~oms/statcvs
StatCVS commands:
java -jar ~/bin/statcvs-xml-0.9.6-full.jar -exclude "**/*.cat:**/*.fits" ../logfile.log ..
java -jar ~/bin/statcvs-xml-0.9.6-full.jar -exclude "**/*.mdl:**/*.gz:**/*.out:**/*.eqn:**/*.ps:**/*.cat:**/*.fits" ../logfile.log ..