The KILOGAS Data Products Pipeline is an automated tool designed to process ALMA CO(2-1) data cubes. It generates a comprehensive set of data products, including moment maps, spectra, uncertainty/upper-limit maps, and final summary panels.

The entire pipeline is controlled and executed through a single main script: run_pipeline.py.

Before running the pipeline, you must configure the primary parameters at the top of the run_pipeline.py script to match your current analysis goals.

Key Parameters:

• version (e.g., 2.0): Defines the output directory version where products will be saved.
• spec_res (e.g., 10 or 30): Selects the spectral resolution (in km/s) of the input data cubes.
• ifu_match (True / False): Determines whether to use IFU-matched cubes (True) or original/nyquist cubes (False).
• clear_save_directory (True / False): If True, completely deletes and resets the output directory before generating new products.

How to Run: Execute the following command in your terminal:

python run_pipeline.py

Once executed, run_pipeline.py runs through the following 5 sequential steps:

1. STEP 1: Smooth and Clip (smooth_and_clip.py)

• Uses the Dame et al. (2011) method to separate noise and create a signal mask.
• Outputs _clipped_cube.fits and _mask_cube.fits. (Note: This is the most computationally intensive and time-consuming step).

2. STEP 2: Moment Map Creation (create_moments.py)

• Generates Moment 0, 1, and 2 maps based on the galaxy's detection status.
• Creates maps in various units (K km/s, Msol/pc², Msol/pix) along with their corresponding error and 3σ upper limit maps.

3. STEP 3: Moment Map Imaging (image_moments.py)

• Renders the generated FITS files into high-quality PNG and PDF images.
• Automatically overlays contour lines, color bars, and synthesized beam sizes.

4. STEP 4: Spectrum Extraction (create_spectrum.py)

• Extracts the CO spectrum from the detected emission area (or within the R50 radius for non-detections).
• Saves the data as a CSV file and plots Brightness Temperature (K) & Flux Density (mJy) vs. Velocity (km/s).

5. STEP 5: Summary Panels (create_summary_panel.py)

• Compiles all individual visual products into a single summary figure per galaxy.
• Detections: 3x3 grid (8 maps + 1 spectrum plot).
• Non-detections: 1x2 grid (Upper limit map + spectrum plot).

If you need to re-run the pipeline for specific galaxies (e.g., for debugging or manual review) without processing the entire dataset, you can override the target list in run_pipeline.py:

# 1. Manually override the target list
targets = ['KGAS10', 'KGAS105', 'KGAS108']

# 2. IMPORTANT: Update the detection lists based on the new targets
detections = [g for g in targets if g in detections]
non_detections = [g for g in targets if g in non_detections]

If the masking and clipping process (Step 1) has already been successfully completed and you only need to regenerate images, moment maps, or spectra, you can significantly reduce execution time by skipping Step 1.

Simply comment out the perform_smooth_and_clip function call near the bottom of run_pipeline.py:

    # print("STEP 1/5: Smooth and Clip")
    # smooth_and_clip.perform_smooth_and_clip(...) # <- Comment this out

The pipeline will bypass the heavy cube processing and directly use the existing _clipped_cube.fits to quickly generate the remaining products.