Telescope Simulation

Telescope simulation is done with the Geant based simulation package AllPix. See the official pages for more detail and installation instructions.

Allpix has the advantage that it comes with a converter as part of the simulation which will format the output to be read by EUTelescope. Hence, the simulated data can be run through the reconstruction chain analogously to raw data.

Basic Features

The code is run (once set-up) in the allpix/tags/AllPix_v1.0 directory. The sub-directories contain the details of simulation specification.

Execution typically is done as follow: > cd [your allpix directory]
> source setup_allpix.sh
> allpix [your macro file]

Steering macros

Steering files are found in the allpix/tags/AllPix_v1.0/macros directory. Here the usual Geant parameters are specified (e.g. beam type, beam energy, # of events) as well as the detector type and positions. As many detectors as you like can be specified so long as there is no overlap. NB the more complicated the simulation, the longer it will take.

Sensor specification

Sensors are specified in the allpix/tags/AllPix_v1.0/models directory, specifically in the pixeldetector.xml file. There are several sensor types already made e.g. FeI4, TimePix and FeI3. Each has a numeric code which is referred to in the steering macro.

Converting to EUTelescope format

After execution there will be a collection of (Mediipix style) output files in the directory specified in the steering macro, e.g. run00001/. The collection is converted as a whole so first the collection must be tar-ed and zip-ed up , e.g. > tar -cvzf run00001.tar.gz ./run00001/*

Next, the whole collection is converted to a single lcio (i.e. EUTelescope readable file) by the share/TelescopeConverter.py file. This is done in python and requires some additional libraries which are found by source-ing the share/setup_pyLCIO.sh file NB this has the side-effect of messing with the AllPix set-up (and vice versa) so the set up files must be re-source-ed.

Encoding

The TelescopeConverter.py contains the details for encoding the AllPix data in an EUTelescope format. The processors of the reconstruction and the data format must match otherwise reconstruction will fail. This can be done either by changing the processor code after conversion, or, and more easily, by carefully setting the conversion encoding.

Before running the conversion make sure the TelescopeConverter.py is set-up to encode the data in the correct format i.e. that which is read by the Clustering (and other) processors of the EUTelescope code. Usually this is 'sparsePixelType'. In addition the usual clustering processors prefer sparsePixelType = 2 (but check these things with reconstruciton code). Therefore, make sure the converter file contains a line like idEncoder_DUT['sparsePixelType'] = 2. For the Telescope planes (i.e. the mimosas) the equivalent line is idEncoder_Telescope['sparsePixelType'] = 1. Also, make sure the event collection has the correct name e.g. event.addCollection( DUTDataColl, 'zsdata_apix' ).

NB The encoding settings can be checked after conversion using the dumpevent of EUTelescope.

Once this is done run the damn code. For example for a DUT with code 200: > python share/TelescopeConverter.py run000001.tar.gz 200

The resulting file is in lcio format. This is equivalent to the converter step of EUTelescope. Check the DUT/mimosa plane id numbers using dumpevent to set the gear file correctly, the simulated data is ready to be clustered.

-- Kenneth Wraight - 2014-05-19

Comments

Topic revision: r1 - 2014-05-19 - KennyWraight
 
This site is powered by the TWiki collaboration platform Powered by PerlCopyright © 2008-2018 by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback