# Licensed under a 3-clause BSD style license - see LICENSE.rst
from __future__ import absolute_import, division, print_function, unicode_literals
import numpy as np
from ..extern import six
from astropy.utils.data import download_file
from astropy.coordinates import Angle, SkyCoord
from astropy.table import Table, Column
from .core import SourceCatalog, SourceCatalogObject
__all__ = [
'SourceCatalogSNRcat',
'SourceCatalogObjectSNRcat',
]
[docs]class SourceCatalogObjectSNRcat(SourceCatalogObject):
"""One source from the SNRcat catalog."""
pass
[docs]class SourceCatalogSNRcat(SourceCatalog):
"""SNRcat supernova remnant catalog.
`SNRcat <http://www.physics.umanitoba.ca/snr/SNRcat/>`__
is a census of high-energy observations of Galactic supernova remnants.
This function downloads the following CSV-format tables
and adds some useful columns and unit information:
* http://www.physics.umanitoba.ca/snr/SNRcat/SNRdownload.php?table=SNR
* http://www.physics.umanitoba.ca/snr/SNRcat/SNRdownload.php?table=OBS
This only represents a subset of the information available in SNRcat,
to get at the rest we would have to scrape their web pages.
* ``table`` (`~astropy.table.Table`) -- SNR info table
* ``obs_table`` (`~astropy.table.Table`) -- High-energy observation info table
Each table has a ``version`` string containing the download date in the ``table.meta`` dictionary.
"""
name = 'snrcat'
description = 'SNRcat supernova remnant catalog.'
source_object_class = SourceCatalogObjectSNRcat
def __init__(self, cache=False):
# TODO: load from gammapy-extra?
# At least optionally?
self.snr_table = _fetch_catalog_snrcat_snr_table(cache=cache)
self.obs_table = _fetch_catalog_snrcat_obs_table(cache=cache)
super(SourceCatalogSNRcat, self).__init__(table=self.snr_table)
def _fetch_catalog_snrcat_snr_table(cache):
url = 'http://www.physics.umanitoba.ca/snr/SNRcat/SNRdownload.php?table=SNR'
filename = download_file(url, cache=cache)
# Note: currently the first line contains this comment, which we skip via `header_start=1`
table = Table.read(filename, format='ascii.csv', header_start=1, delimiter=';')
table.meta['url'] = url
table.meta['version'] = _snrcat_parse_download_date(filename)
# TODO: doesn't work properly ... don't call for now.
# _snrcat_fix_na(table)
table.rename_column('G', 'Source_Name')
table.rename_column('J2000_ra (hh:mm:ss)', 'RAJ2000_str')
table.rename_column('J2000_dec (dd:mm:ss)', 'DEJ2000_str')
data = Angle(table['RAJ2000_str'], unit='hour').deg
index = table.index_column('RAJ2000_str') + 1
table.add_column(Column(data=data, name='RAJ2000', unit='deg'), index=index)
data = Angle(table['DEJ2000_str'], unit='deg').deg
index = table.index_column('DEJ2000_str') + 1
table.add_column(Column(data=data, name='DEJ2000', unit='deg'), index=index)
radec = SkyCoord(table['RAJ2000'], table['DEJ2000'], unit='deg')
galactic = radec.galactic
table.add_column(Column(data=galactic.l.deg, name='GLON', unit='deg'))
table.add_column(Column(data=galactic.b.deg, name='GLAT', unit='deg'))
table.rename_column('age_min (yr)', 'age_min')
table['age_min'].unit = 'year'
table.rename_column('age_max (yr)', 'age_max')
table['age_max'].unit = 'year'
distance = np.mean([table['age_min'], table['age_max']], axis=0)
index = table.index_column('age_max') + 1
table.add_column(Column(distance, name='age', unit='year'), index=index)
table.rename_column('distance_min (kpc)', 'distance_min')
table['distance_min'].unit = 'kpc'
table.rename_column('distance_max (kpc)', 'distance_max')
table['distance_max'].unit = 'kpc'
distance = np.mean([table['distance_min'], table['distance_max']], axis=0)
index = table.index_column('distance_max') + 1
table.add_column(Column(distance, name='distance', unit='kpc'), index=index)
table.rename_column('size_radio', 'diameter_radio_str')
diameter_radio_mean = _snrcat_parse_diameter(table['diameter_radio_str'])
index = table.index_column('diameter_radio_str') + 1
table.add_column(Column(diameter_radio_mean, name='diameter_radio_mean', unit='arcmin'), index=index)
table.rename_column('size_X', 'diameter_xray_str')
diameter_xray_mean = _snrcat_parse_diameter(table['diameter_xray_str'])
index = table.index_column('diameter_xray_str') + 1
table.add_column(Column(diameter_xray_mean, name='diameter_xray_mean', unit='arcmin'), index=index)
table.rename_column('size_coarse (arcmin)', 'diameter_mean')
table['diameter_mean'].unit = 'arcmin'
table.rename_column('size_imprecise', 'diameter_mean_is_imprecise')
return table
def _fetch_catalog_snrcat_obs_table(cache):
url = 'http://www.physics.umanitoba.ca/snr/SNRcat/SNRdownload.php?table=OBS'
filename = download_file(url, cache=cache)
# Note: currently the first line contains this comment, which we skip via `header_start=1`
table = Table.read(filename, format='ascii.csv', header_start=1, delimiter=';')
table.meta['url'] = url
table.meta['version'] = _snrcat_parse_download_date(filename)
# TODO: doesn't work properly ... don't call for now.
# _snrcat_fix_na(table)
return table
def _snrcat_fix_na(table):
"""Fix N/A entries in string columns in SNRcat."""
for colname in table.colnames:
if isinstance(table[colname][0], six.text_type):
mask1 = (table[colname] == 'N / A')
mask2 = (table[colname] == 'N/A')
table[colname].mask = mask1 | mask2
table[colname].fill_value = ''
def _snrcat_parse_download_date(filename):
text = open(filename).readline()
# Format: "This file was downloaded on 2015-06-07T03:39:53 CDT ..."
tokens = text.split()
date = tokens[5]
return date[:10]
def _snrcat_parse_diameter(text_col):
"""Parse SNRcat diameter string column"""
d_means = []
for text in text_col:
try:
# Parse this text field:
if 'x' in text:
a, b = text.split('x')
d_major = Angle(a).arcmin
d_minor = Angle(b).arcmin
else:
d_major = Angle(text).arcmin
d_minor = d_major
d_mean = _snr_mean_diameter(d_major, d_minor)
except:
# print('Parsing error:', text)
d_mean = np.nan
d_means.append(d_mean)
return d_means
def _snr_mean_diameter(d_major, d_minor):
"""Compute geometric mean diameter (preserves area)"""
diameter = np.sqrt(d_major * d_minor)
# If no `d_minor` is given, use `d_major` as mean radius
with np.errstate(invalid='ignore'):
diameter = np.where(d_minor > 0, diameter, d_major)
return diameter