some functions moved to utils

README.md

@@ -18,7 +18,7 @@ Features
 
 - Convolution-based adaptive S/N approach for detecting absorbers in QSO spectra.
 - Gaussian fitting for accurate measurement of absorber properties (such as EW, line widths, and centers).
-- Parallel processing for efficient computation on a large number of spectra.
+- Parallel processing using multiprocessing for efficient computation on a large number of spectra.
 
 Documentation
 -------------
@@ -76,9 +76,9 @@ Useful notes:
 Parallel mode can be memory-intensive if the input FITS file is large in size. As the code accesses the FITS file to read QSO spectra when running in parallel, it can become a bottleneck for memory, and the code may fail. Currently, I suggest the following:
 
 - **Divide your file into smaller chunks:** Split the FITS file into several smaller files, each containing approximately `N` spectra. Then run the code on these smaller files.
-           
+
 - **Use a rule of thumb for file size:** Ensure that the size of each individual file is no larger than `total_memory/ncpu` of your node or system. Based on this idea you can decide your `N`. I would suggest `N = 1000`.
-           
+
 - **Merge results at the end:** After processing, you can merge your results.
 
 In order to decide the right size of the FITS file, consider the total available memory and the number of CPUs in your system.

docs/index.rst

@@ -12,7 +12,7 @@ Features
 
 - Convolution-based adaptive S/N approach for detecting absorbers in QSO spectra.
 - Gaussian fitting for accurate measurement of absorber properties (such as EW, line widths, and centers).
-- Parallel processing for efficient computation on a large number of spectra.
+- Parallel processing using multiprocessing for efficient computation on a large number of spectra.
 
 .. toctree::
    :maxdepth: 1

qsoabsfind/parallel_convolution.py

@@ -5,26 +5,12 @@
 import numpy as np
 import argparse
 import time
+import os
 from multiprocessing import Pool
 from .absfinder import read_single_spectrum_and_find_absorber
 from .io import save_results_to_fits
-import re
-import os
-import pkg_resources
 from .config import load_constants
-from .utils import read_nqso_from_header
-
-def get_package_versions():
-    """
-    Get the versions of qsoabsfind and other relevant packages.
-
-    Returns:
-        dict: A dictionary containing the versions of the packages.
-    """
-    packages = ['qsoabsfind', 'numpy', 'astropy', 'scipy', 'numba', 'matplotlib']
-    versions = {pkg: pkg_resources.get_distribution(pkg).version for pkg in packages}
-    return versions
-
+from .utils import read_nqso_from_header, get_package_versions, parse_qso_sequence
 
 def run_convolution_method_absorber_finder_QSO_spectra(fits_file, spec_index, absorber, kwargs):
     """
@@ -42,34 +28,6 @@ def run_convolution_method_absorber_finder_QSO_spectra(fits_file, spec_index, ab
     """
     return read_single_spectrum_and_find_absorber(fits_file, spec_index, absorber, **kwargs)
 
-def parse_qso_sequence(qso_sequence):
-    """
-    Parse a bash-like sequence or a single integer to generate QSO indices.
-
-    Args:
-        qso_sequence (str or int): Bash-like sequence (e.g., '1-1000', '1-1000:10') or an integer.
-
-    Returns:
-        numpy.array: Array of QSO indices.
-    """
-    if isinstance(qso_sequence, int):
-        return np.arange(qso_sequence)
-
-    # Handle string input
-    if isinstance(qso_sequence, str):
-        if qso_sequence.isdigit():
-            return np.arange(int(qso_sequence))
-
-        match = re.match(r"(\d+)-(\d+)(?::(\d+))?", qso_sequence)
-        if match:
-            start, end, step = match.groups()
-            start, end = int(start), int(end)
-            step = int(step) if step else 1
-            return np.arange(start, end + 1, step)
-
-    # If none of the conditions matched, raise an error
-    raise ValueError(f"Invalid QSO sequence format: '{qso_sequence}'. Use 'start-end[:step]' or an integer.")
-
 def parallel_convolution_method_absorber_finder_QSO_spectra(fits_file, spec_indices, absorber, n_jobs, **kwargs):
     """
     Run convolution_method_absorber_finder_in_QSO_spectra in parallel using

qsoabsfind/utils.py

@@ -10,6 +10,8 @@
 import os
 from astropy.io import fits
 from astropy.table import Table
+import re
+import pkg_resources
 
 # Configure logging
 #logging.basicConfig(level=logging.DEBUG, format='%(asctime)s - %(levelname)s - %(message)s')
@@ -20,6 +22,46 @@
 constants = load_constants()
 lines, amplitude_dict, speed_of_light = constants.lines, constants.amplitude_dict, constants.speed_of_light
 
+def get_package_versions():
+    """
+    Get the versions of qsoabsfind and other relevant packages.
+
+    Returns:
+        dict: A dictionary containing the versions of the packages.
+    """
+    packages = ['qsoabsfind', 'numpy', 'astropy', 'scipy', 'numba', 'matplotlib']
+    versions = {pkg: pkg_resources.get_distribution(pkg).version for pkg in packages}
+    return versions
+
+def parse_qso_sequence(qso_sequence):
+    """
+    Parse a bash-like sequence or a single integer to generate QSO indices.
+
+    Args:
+        qso_sequence (str or int): Bash-like sequence (e.g., '1-1000', '1-1000:10') or an integer.
+
+    Returns:
+        numpy.array: Array of QSO indices.
+    """
+    if isinstance(qso_sequence, int):
+        return np.arange(qso_sequence)
+
+    # Handle string input
+    if isinstance(qso_sequence, str):
+        if qso_sequence.isdigit():
+            return np.arange(int(qso_sequence))
+
+        match = re.match(r"(\d+)-(\d+)(?::(\d+))?", qso_sequence)
+        if match:
+            start, end, step = match.groups()
+            start, end = int(start), int(end)
+            step = int(step) if step else 1
+            return np.arange(start, end + 1, step)
+
+    # If none of the conditions matched, raise an error
+    raise ValueError(f"Invalid QSO sequence format: '{qso_sequence}'. Use 'start-end[:step]' or an integer.")
+
+
 def elapsed(start, msg):
     """
     Prints the elapsed time since `start`.