Dataset pipeline for H-21cm cube

For the production of epoch of reionisation (EoR) hydrogen 21cm (H-21cm) observational .fits cubes. This pipeline utilises a modified/forked version of the python implementation of the 21cmfast library.

Installing

Conda Environment

In order to use this package, we have provided a conda environment configuration file (environment.yaml). You can use this to create your conda virtual environment with which to run the pipeline to produce the H-21cm cube. In order to set up the environment on the command-line:

conda env create --file environment.yaml

This will create an environment named 'sdc3_h21cm_generator' the envs directory within conda's base directory. In order to create the virtual environment in a custom path (e.g. ./venv):

conda env create -p ./venv --file environment.yaml

You can also create a custom-named environment via:

conda env create -n my_name --file environment.yaml

Singularity Image

Coming soon.

Operation

Parameter file

Using the pipeline requires the configuration of a .toml file containing all of the required parameters needed for an execution of the pipeline. The full path to this configuration file constitutes the sole command-line argument for the execution of the pipeline.

The parameter file is divided into the following sections:

• outputs: Directories and file naming
• field: Observational coordinate system
• correlator: Frequency setup
• astro_params: 21cmfast EoR realisation parameters
• user_params: Various user and 21cmfast configuration parameters
• flags: Flags to pass to 21cmfast

An example parameter file is provided in this repository in files/example_params.toml with annotations in the form of comments for each individual parameter.

Conda environment

After setting up the virtual environment, it can be activated via:

conda activate sdc3_h21cm_generator

Once within the virtual environment, the pipeline can be executed according to the parameters within the parameter file via:

python /path/to/sdc3_training_datasets/eor_h21cm.py parameter_file.toml

Singularity image

Coming soon.

Dependencies

Please see the file, environment.yaml, for a full list of dependencies.