Toolchain for analysing data of the icecube experiment.

• data_handler tools for managing data sets
• icetray_module modules relying on icetray
• job_handler are scripts for batch handling/organisation of data sets
• plotting collection of plotting tools

This script can be used to create DAGman jobs from a config file. The templates for all files created during the process can be found in the resources folder.

It is intended to create jobs processing folders (with sub-folders) of data one file at a time, whereas each file should be processed in one job.

Provide a init file like the example config in resources/config.ini.

Run ./bin/create_dagman_job /path/to/config.ini.

This will create a folder named after the provided job name, with a copy of the init file for reproducability. Inside this folder folders for each dataset are created. A dataset number is the combination of the particle_type and the systematics value.

Within the job section number of jobs to submit per interval and max jobs to submit in total are given, as well as the overall DAGman job name. The DAGman job limitations are stored in a dagman.config file.

data provides information about the given input filename syntax for all files in settings.inputfile. As well as a desired output syntax using elements from the input syntax. Elements not found in the input syntax are used literaly.

All values provided in the settings section are passed as options to the process.executable, thus creating a list of all options for DAGman in a options.dag file.

The process section is used to define the used software environment, keep track of the used code and its commit version, as well as the actual program to execute. This results in a OneJob.submit file.

A job created with this script can be started via

condor_submit_dag -config dagman.config -notification Complete options.dag