PACS - PCSS

Generate User Manual from XML as PDF

• cd vs-pcss-src>/pacs/develop/pcss/

• checkout the doc-files from the herschel.pacs.doc-files

• cd to <cvs-pcss-src>/pacs/develop/pcss/herschel/pacs/doc-files

• make you changes in the *.xml files in help/um/xml
• make sure the following env variables are set:
  • $HCSS_DIR pointing to your hcss distribution
  • $PCSS_OUT pointing to your development out directory (in this example $HOME/out)
• make a file jake.root

  •       hcss.jake.outdir = $PCSS_OUT

• run the make_um.local from the doc-files directory

  •          help/bin/make_um.local pdf 

• In the moment you might get an error : xception in thread "main" java.lang.OutOfMemoryError: Java heap space

• If you get no error
  • check the result in your web browser pointing to $HOME/out/pcss/doc/help/um/html/um.html

• if you get an error : exception in thread "main" java.lang.OutOfMemoryError: Java heap space

  • you find the $HOME/out/doc/help/um/pdf/um.fo file which will need to be converted (with the help of FOP http://xmlgraphics.apache.org/fop/download.html#binary) to PDF

  •            java -Xmx3g herschel.ia.document.tools.Fo \
                  -fo ${LOCATION_PDF}/${DOC_NAME}.fo ${LOCATION_PDF}/${DOC_NAME}.pdf 

  • check the result in your web browser pointing to $HOME/out/pcss/doc/help/um/html/um.html

Attention: for the conversion to work properly you have to install a version of dbdoclet.jar. This software comes with its own version of jfreechart. If the dbdoclet.jar is in the classpath before the hcss external libraries, also hcss will use the jfreechart version of dbdoclet. This leads to problems at least with PlotXY and PacsQla. So make sure dbdoclet.jar is added at the end of your classpath!

Usage of TestExecutionBrowser just to get an overview

In JIDE :

from herschel.access.util import *
t=TestExecutionBrowser()

Size of level 0.5 data structures in PCSS

Typical Spectrometer observation after decompression :

One second of Spectrometer data (default mode/3 raw channels) :

Basic science data + reduction add on (error, wavelength)

Detectors : 26 * 18 Ramp length : 64 Readout frequency : 256 Hz NrSubRamps : 8 All data as doubles : 8 Byte

Readouts : 26 * 18 * 256/64 * 8 * 8 = 120 KB Wavelength : same as above : 120 KB Error : same as above : 120 KB

Raw data

Readout frequency : 256 Hz All data as doubles : 8 Byte Nr raw channels : 3

Raw data : 256 * 8 * 3 = 6 KB

DecMec data:

Readout frequency : 256 Hz All data as int : 4 Byte DecMec entries : 13

256 * 4 * 13 = 13KB

==> ~ 380 KB per second

One hour :

==> ~ 1,3 GB per hour

Observation duration <= 3hour (there may be exceptions, but then we reduce the number of user)

==> ~ 4 GB per observation

Before and after reduction step

==> ~ two times before and after processing step 8 GB /person

Product Access Layer

The Product Access Layer (PAL) is now available in PCSS. We encourage everybody to use it. Basic documentation you find under :

ftp://ftp.rssd.esa.int/pub/HERSCHEL/csdt/releases/doc/ia/document/urm/html/ch02.html

A typical example for Frames is :

storage=ProductStorage()
storage.register(SimplePool.getInstance()) 
reference=storage.save(frames)
print reference.urn

from herschel.pacs.signal import *
reference = storage.load("urn:simple.default:herschel.pacs.signal.Frames:1")

#get back your saved product
frame = reference.getProduct()

To use the GUI Browser to find your Products you may use :

from herschel.ia.pal.browser import ProductBrowser
from herschel.ia.pal.browser import ProductBrowserDialog

p=ProductBrowser(storage)
d=ProductBrowserDialog(p)

To get out the Data you need to move it in the Basket (marking in the GUI). Then you may extract it from the basket :

prod=p.getJideBasket().toArray()[0].getProduct()