Mode Configuration

Version Update (v0.6.0)

Since version 0.6.0 you do not have to manually register your mode, but the mode is automatically registered dependent on where you defined it. For example: A mode defined under vibe.analysis_validation_modes.physics is automatically associated with the DPGroup.physics.

To get your mode to run you have to define a config with the information about the datasets this mode should run on. This is handled via a toml file with the same name as the python file in which you defined your mode. The file has to be placed in the same directory as the python file containing your mode. For example: You created the mode MyExampleMode under vibe.analysis_validation_modes.physics.my_example_mode.py, then the path to your config is vibe.analysis_validation_modes.physics.my_example_mode.toml.

The config contains several blocks of information:

1. Information about the datasets to run on: In this code block all datasets the mode will run on have to be defined with unique identifiers. The code block is created by the keyword [dataset_dict] and each identifier within this block must start with [dataset_dict.your_identifier]. Each dataset has an lpn which is the only required parameter to be set and can either be a path on the grid or a local path. In the latter case, one has to add the additional argument offline = true. Furthermore, for each dataset it is possible to define individual kwargs that are used in the processing of the datasets. An example dataset_dict could look like this:

[dataset_dict]
    [dataset_dict.release-08-00-00]
    lpn = "/path/to/lpn/on/grid/mdst"

    [dataset_dict.release-08-00-00]
    lpn = "/path/to/another/lpn/on/grid/mdst"
    kwargs = {"test": 1}

    [dataset_dict.prerelease-09-00-00c]
    lpn = "/path/to/local/files/on/kekcc/mdst"
    offline = true

Tip

Setting local datasets that are run offline is especially useful for testing either the mode works like intended or not!

Note

As offline=false is the default, this does not have to be explicitly specified for datasets that are run on the grid with gbasf2.

2. Additional gbasf2 parameters (optional): It is possible to set the gbasf2 CPU time or give additional gbasf2 parameters in the [gbasf2_settings] block. For example:

[gbasf2_settings]
    gbasf2_cpu_time = 1000
    gbasf2_additional_params = "--banned_site=LCG.Torino.it,LCG.ULAKBIM.tr"

Note

Note that these settings are applied to all datasets run on with gbasf2.