The data and code for : Sutra: An integrated framework for identification and characterization of filaments in the interstellar medium

Sutra is a machine learning based framework for the identification and characterization of interstellar filaments in molecular cloud observations. The package combines U-Net based crest extraction with beam-scale physical filtering using Plummer profile analysis. Sutra provides beam-level filament properties including width, contrast, linear mass density, and profile shape parameters. The software is available as a Python package with both CLI and Streamlit-based interfaces for large-scale filament analysis. Installation git clone https://git.sac.gov.in/destiny_doom/sutra.git cd sutra pip install . Note : we recommend creating a new conda environment before installation and make sure that current `pip` is coming from conda environment conda create -n conda activate which pip verify installation using following command: sutraWeb ## Web Application: The streamlit based web application runs with the following command from anywhere in the terminal sutraWeb ## Command Line Application sutraTracer -c -s -p #### CLI options ``` options: -h, --help show this help message and exit -c INPUT_FILE, --cd_file INPUT_FILE Path to the input Column density map file. -s SKL_OUTPUT_FILE, --skl_output SKL_OUTPUT_FILE Path to the output Skeleton map. -p MODEL_OUTPUT_FILE, --model_output MODEL_OUTPUT_FILE Path to the output Skeleton map. -t THRESHOLD, --threshold THRESHOLD Threshold to convert model output to skeleton map -m MODEL, --model MODEL Dictionay key-name storing the path to trained model weights ``` ## Using Sutra as Python package ### Import Sutra ```python from sutra.tracer.predictor import filamentIdentifier as FID from sutra.measurement import local_field as LF ``` ### Create Crest probability map using trained model ```python monr2_crop = fits.open(' ') ``` ```python prob_map = FID('HGBS').predict(monr2_crop[0], window_overlap_frac = 0.95, batch_size = None, n_jobs = 1) ``` ```python meta_info = { 'distance' : 830, # 140 pc for Taurus , 260 for musca 'beam' : 36.4 , # arcsec 'radial-cutoff' : 0.6 , } ``` ### Skeletonize the probability map ```python from sutra.profilerV2.prob2skl import run_skel th_max = 0.45 beam_size = 12 skel = run_skel(prob_map, th_max = th_max, bkg_mask=None, beam_size=12, prune = True, convolve_map=True) ``` ### Use Skeleton for measuring filament properties ```python from sutra.profilerV2.radprof import RadProf monr2aprof = RadProf(monr2_crop[0], skel, meta_info=meta_info) monr2aprof.tangents(ks = 3, stride=2) # create radial cuts monr2aprof.reorder() # give location based ordering to each radial profiles monr2aprof.spline_smooth() # smoothens the skeleton spine using Bezier curves monr2aprof.cut_off_points(alpha_pc=1.5) #cutoff-distance in parsec monr2aprof.create_rad_profile_single_thread() # Extract all filament profiles and store in monr2Prof.beam_dict['beam_elements'] ``` ```python monr2aprof.group_profiles() # Group profiles together in beam elements monr2aprof.get_all_beam_props() # Compute beam level properties ``` ### Properties Map ```python monr2aprof.plot_props(sizescale=0.5, sizeby='Wbg', show_filid=True, red_chi_filter=5, colorby='Nfil') ``` ### Plot a selected Filament ```python monr2aprof.plot_filament(49, sizescale=10, red_chi_filter=5, sizeby='Wbg', show_beamid=True) ``` ### Inspect the Radial profile fit of selected beam ```python monr2aprof.beam_dict['beam_elements'][374].plot_plummer() ``` ng) ### Access the Properties map data-table ```python monr2aprof.beamProps ```