HiFAST Command-Line Tools: General Usage#

This page describes the general usage patterns for all hifast subcommands. There are two main ways to execute a command, depending on whether you are processing a single file or multiple files in parallel.

1. Single-File Processing (hifast.<subcommand>)#

The most direct way to run a hifast module is by using a dot (.) between hifast and the subcommand name. This syntax is intended for processing a single input file.

Syntax:

python -m hifast.<subcommand> [input_file] [options]

Example:

python -m hifast.sep my_data_M01.fits --frange 1400 1440

2. Parallel Multi-File Processing (hifast <subcommand>)#

To make processing large datasets easier, hifast provides a convenient wrapper to run a command on multiple files in parallel. This method uses a space between hifast and the subcommand name.

Under the hood, this wrapper uses xargs to execute the single-file command (hifast.<subcommand>) for each input file.

Syntax:

python -m hifast <subcommand> -p <N> [input_file_1] [input_file_2] ... [options]

  • The -p <N> or –nproc <N> flag is crucial here; it specifies that N processes should be run in parallel.

Example:

python -m hifast sep -p 4 M01.fits M02.fits M03.fits M04.fits --frange 1400 1440

Note

Most processing subcommands (like sep, bld, rfi, etc.) support this parallel wrapper.

Other Parallel Processing Methods#

For more complex workflows, such as chaining multiple commands together to run on a set of files, the hifast.sh script is available. Please see the Parallel page for detailed examples of its use.

Getting Help#

To see a comprehensive list of all available parameters for any subcommand, you can use the -h or –help flag. This works for both syntaxes:

# Get help for a specific module's parameters
python -m hifast.sep -h | more

# Get help for the parallel wrapper itself
python -m hifast sep -h | more

Inspecting Output Files#

The intermediate files generated by hifast are typically in the HDF5 format (.hdf5). You can inspect these files from the command line to see their contents and metadata, or visualize them to see the spectra.

Command-Line Inspection

To see the commands and parameters used to create a file, use h5dump:

h5dump -g /Header /path/to/your_file.hdf5

To see a detailed, recursive listing of the contents of an HDF5 file, use h5ls:

h5ls -r /path/to/your_file.hdf5

Visual Inspection (Waterfall Plots)

A common way to inspect the data is to create a “waterfall plot” (a 2D plot of frequency vs. time). You can do this using the dedicated hifast.waterfall subcommand.

For detailed instructions on how to generate these plots and view them with tools like CARTA, please see the Waterfall Plot page.