Parameterized jobs - ALMA RASCIL vs. ALMA CASA tests¶
This documentation offers CASA vs. RASCIL scripts on ALMA datasets, for results comparison:
Three ALMA datasets are already uploaded under LFN, these are: HLTau_Band6, HLTau_Band7 and Mira_Band6. ALMA datasets are stored under LFN as: %s_CalibratedData.tgz, where %s is the name of the parameter:
Parameters = {"HLTau_Band6","HLTau_Band7","Mira_Band6"};
see the parameters in the .jdl
ALMA RASCIL¶
Scripts:
Folder alma_rascil contains all scripts. Download files and submit to IRIS three jobs, by using the below command:
$ dirac-wms-job-submit alma_rascil.jdl
- More parameters can be added to the Parameters={“HLTau_Band6”,,,,} list above, along with the ALMA datasets and their own configuration files. See example config files in configfiles.tar.gz. The steps, to add another dataset to be processed, are:
download ALMA dataset and store it under LFN with the name %s_CalibratedData.tgz (%s being one of the parameters, eg. HLTau_Band6)
create a configuration file for this dataset and add it to the configfiles.tar.gz
add the new parameter to the list of parameters in the .jdl file, Parameters = {“HLTau_Band6”,”HLTau_Band7”,”Mira_Band6”};
ALMA RASCIL tests are using a RASCIL container build from a recipe to enable the usage of config files. An example of recipe file that can be used is:
recipe file that uses an already pulled local RASCIL container: RASCIL-fullN.img $ cat SingRascilCasa.recipe bootstrap: localimage from: /home/<user>/RASCIL-fullN.img %post cd /var/lib git clone https://github.com/casacore/casacore apt-get update && \ apt-get -y install sudo sudo apt-get -y install build-essential cmake gfortran g++ libncurses5-dev \ libreadline-dev flex bison libblas-dev liblapacke-dev libcfitsio-dev \ wcslib-dev libfftw3-dev sudo apt-get -y install libhdf5-serial-dev python-numpy \ libboost-python-dev libpython3.7-dev libpython2.7-dev cd casacore mkdir build cd build cmake .. make make install building the new container from recipe uses the command: $ sudo singularity build RASCIL-full.img SingRascilCasa.recipe
ALMA CASA¶
Scripts:
Folder alma_casa contains all scripts. Download files and submit to IRIS three jobs, by using the below command:
$ dirac-wms-job-submit alma_casa.jdl
ALMA CASA scripts have been downloaded and their name has been changed to %s_Imaging.py, where %s is the name of the parameter (because of use of parametrized jobs). The files will be then archived into imgfiles.tar. These scripts are using the ALMA datasets %s_CalibratedData.tgz mentioned above.
RESULTS HLTau_Band6 DATASET¶
These are the results for HLTau_Band6 DATASET ALMA_RASCIL and ALMA_CASA Two fits images (RASCIL vs CASA outputs) converted to png are being shown below:
