Frequently asked questions concerning the SPI analysis software

last update of this page: 7.10.2005

In case you have a problem with the SPI analysis, it might well be that this question was raised already in the past. Find here some standard questions and answers. Please have a look on this list before contacting the GSFC/ISDC SPI team. In case it is a more general question, you might also find an answer on the ISDC FAQs page. Thanks!

Contents

Is it possible to use the Observation Simulator (OSIM) for SPI ?
OSIM is not including SPI for the moment. But you can use the script spi_osim 1.0 (which includes gensky 11.0, os_pdefgen 4.1, spisimprep 3.0, and spiskycnv 16.0) for SPI simulation. Read the SPI simulation user manual to see how this works.
Where I can find the SPI related documents?
Use this link
Where do I find the PSD documentation?
Use Jürgen's documentation for description of the PSD itself and for a general PSD Software Description.
Look at Jürgen's SPI page for documentation of the individual PSD executables.
How do I update an existing component in my ISDC/SPI installation?
- Download the new component from the ISDC software page (be patient, loading the page takes a while).
- Put it in the directory where you want to install it (e.g. below your $ISDC_ENV/ directory).
- unpack it with e.g. gzip -dc new_component.tar.gz | tar xf -
- $ISDC_ENV/bin/ac_stuff/configure
- to install it only in this directory: make
- to install it globally (you need write permission under $ISDC_ENV) : make global_install
- rehash
How do I know the correct dependencies for my software?
To find out about the dependencies of one program:
- go to http://isdc.unige.ch/cgi-bin/cgiwrap/isdc_lib/cgi-bin/info_component.cgi (or go to ISDC homepage, then click on "Config. Mgt", then on the left on "Status of deliveries", then in the page on "Informations on one particular component").
- enter name of component and click on "Component search"
- in the table which is displayed you can click on "Dependencies" to see which components are needed/were used to install the program.
I made a minor mistake in the delivery note. Is there an easy way to re-deliver?
In case you want to re-deliver my_component version Version_Number:
- go to http://isdc.unige.ch/index.cgi?Configmgt+sw_deliv
- type in the name of the component: my_component
- type in the version number of the delivery you want to repeat: Version_Number
- click on "Submit delivery Form"
- Now the last delivery form is displayed.
- Update the form
- click on "Submit delivery Form"
- check whether or not everything is okay
- click on "Confirm Delivery Form"
Is there an exact set of rules for parameter files?
Take care of the following:
- look also at this answer and some explanation of parameter file structure
The parameter file I use for an analysis component/script is not working properly. Where can I find a good one?
In principle the parameter files coming with the software delivery should work (see $ISDC_ENV/pfiles). You can find parameter files for single components/scripts of the SPI software, filled with meaningful values on the Programs & Parfiles page
How can I use a variable in a parameter ?
Put a $ and curly brackets around the variable name:
- setenv MY_VARIABLE /my/value
- use it in the parameter file like
  myparameter,s,h,"${MY_VARIABLE}/whateverfile",,,""
Can I use the output catalogue of SPIROS as an input catalogue (for SPIROS and/or SPISKYMAX) ?
Not directly. Do the following:
- open the SPIROS output catalogue with fv
- click on 'Table' of SPI.-SRCL-RES
- select the sources you want to use by setting SEL_FLAG (third but last column) to 1
- click on 'Header' of SPI.-SRCL-RES
- Change EXTNAME to 'SPI.-SRCL-CAT'
- Save the changes (Alt+S)
Where do I find the meaning of an error number?
The ISDC Error Codes can be found on the ISDC error page. As a developer you might want to check the SPI error code ranges to see which range can be occupied by your software. The error code should also be described in the help text of the component: type my_component --help to see the error code description of my_component (e.g. gti_data_gaps --help)
og_create crashes with error -2005
You have to chose a non-existing obs branch and you have to have write permission on the obs/ directory.
Right at the beginning of my analysis it crashes with status -3003, saying "Parameter initialization error"
Your PFILES variable is not set to a directory where the parameter files for your analysis program can be found. Use e.g. setenv PFILES $ISDC_ENV/pfiles
or setenv PFILES where/you/have/your/parameter/directory/
Repeating part of the analysis, I always get error -2004, even though I set the clobber to 'yes'
Overwriting existing files is sometimes not possible. Remove those files first or chose different output file names. Also make sure that these files are not attached to your observation group (in case you use one). When using the script spi_science_analysis or blc_sa you should not have this problem (when you always use the same observation group)
Running spi_gti_creation the error -1261 occurs, which is DAL_BAD_TFORM.
Probably you were asking for different kinds of GTI which are not accessible in the data set, like "ATTITUDE DATA_GAPS". The data set might be incomplete. Try applying the following spi_gti_creation parameter:
use_gtis,s,ql,"DATA_GAPS",,,"GTI names"
Running a script I get an error "Parameter named XYZ is not a valid parameter"
There seems to be a mismatch between the parameters which are defined for the script and the parameters of a single component, called by the script. Check if the parameter names are the same in both parameter files.
Example: You use the script spi_science_analysis to run the analysis, and the component spi_gain_corr fails with "Parameter named inSWG is not a valid parameter". Then check if the parameter is inside your spi_gain_corr parameter file:
grep inSWG $PFILES/spi_gain_corr.par
Running spi_osim it always crashes with Error -2005 right in the beginning
Probably you forgot to create a subdirectory spi/
Please follow the instructions in the SPI_OSIM user manual
Running a script I get a warning which tells me that "The internal data structures have been changed"
Your software environment setup is probably out of date. Please follow the setup advice (GSFC) or the setup advice (ISDC) and change your .tcshrc, if necessary.
How can I do model-fitting of SPIROS spectra?
You can use XSPEC on your SPIROS output spectra. You also have to have the ftools installed. For a step-by-step description look at the XSPEC/SPIROS spectra page.
How do I define a good time interval?
For a step-by-step description look at this page.
How do I analyse a GRB ?
For a step-by-step description look at this page.
How do I prepare a set of IRFs for delivery ?
For a step-by-step description look at this page.
How do I analyse a very large SPI data set?
Data sets which exceed several hundred pointings are better analysed in subsets and then added together after the binning of the data. For a step-by-step description look at this page.
What is the SPIHIST problem in OSA 3.0?
There is a problem in the SPI analysis software in OSA 3.0. Read more about it!
How do I update my OSA 3.0 installation for SPI?
Some important tools have been updated since OSA 3.0 software release. Here is a step-by-step instruction how to do the updates.
How do I copy a subset of INTEGRAL data?
In some cases you might want to copy a subset of data from the ISDC. It is not necessary to copy all the four directories (scw, aux, ic, idx). Here is a step-by-step how to copy an INTEGRAL data subset.
How do I convert SPIROS count rates into fluxes?
SPIROS count rates are nearly photons/cm**2/sec (as given below), especially at low energies. The conversion factors, based on Crab observations in revolution 44, are:
f[20 - 40 keV]: 1 Crab = 0.1788 ph/cm**2/sec = 7.9 e-9 erg/cm**2/sec
f[40 - 60 keV]: 1 Crab = 0.0688 ph/cm**2/sec = 5.4 e-9 erg/cm**2/sec
f[60 - 100 keV]: 1 Crab = 0.0540 ph/cm**2/sec = 6.6 e-9 erg/cm**2/sec
f[100 - 200 keV]: 1 Crab = 0.0370 ph/cm**2/sec = 8.2 e-9 erg/cm**2/sec
f[200 - 500 keV]: 1 Crab = 0.0200 ph/cm**2/sec = 9.8 e-9 erg/cm**2/sec
f[500 -1000 keV]: 1 Crab = 0.0063 ph/cm**2/sec = 7.0 e-9 erg/cm**2/sec
You find the fluxes of your analysis in the SPIROS logfile and in the output catalog (SPI.-SRCL-RES) given in ph/cm**2/sec
How to extract spectra from an ISGRI mosaic image?
There are several ways to extract spectra from ISGRI data. I do not claim that the one I propose here is the best or the most accurate. Use it wisely. Here is the step-by-step description how to extract ISGRI spectra from a mosaic image.
When do INTEGRAL data become public?
INTEGRAL data become public one year after they have been ingested into the archive. This is roughly 13 to 14 months after the data have been taken. But in case there were processing problems, it could also take much longer. The information about which revolutions become public on what date can be found under the Public Data Release Page at the ISDC.
When did SPI annealing take place?
The 19 SPI detectors need from time to time (twice a year or so) to be heated to about 100^oC and then cooled again to their nominal temperature of 85^oK (-188^oC). This procedure is called SPI annealing and is needed to `repair' the crystal structure of the 19 Germanium detectors that gets degraded by the space environment (cosmic rays, solar wind, etc.). The full procedure takes about two weeks and the instrument cannot be used during that period.
Periods of past SPI annealings:
1st annealing started February 6, 2003, revolution 38-41
2nd annealing July 18-30, 2003, revolution 93-96
3rd annealing after the big solar flare in November 2003, revolution 132-136
4th annealing started June 20, 2004, revolution 206-209
5th annealing started with revolution 278 on January 22, 2005 and ended in revolution 280 on January 29, 2005
6th annealing
7th annealing

You did not find the answer to your question? Even not on the ISDC FAQs page? Then send your question to the INTEGRAL help desk