def fetch_sdss_filter(fname, data_home=None, download_if_missing=True):
"""Loader for SDSS Filter profiles
Parameters
----------
fname : str
filter name: must be one of 'ugriz'
data_home : optional, default=None
Specify another download and cache folder for the datasets. By default
all scikit learn data is stored in '~/astroML_data' subfolders.
download_if_missing : optional, default=True
If False, raise a IOError if the data is not locally available
instead of trying to download the data from the source site.
Returns
-------
data : ndarray
data is an array of shape (5, Nlam)
first row: wavelength in angstroms
second row: sensitivity to point source, airmass 1.3
third row: sensitivity to extended source, airmass 1.3
fourth row: sensitivity to extended source, airmass 0.0
fifth row: assumed atmospheric extinction, airmass 1.0
"""
if fname not in 'ugriz':
raise ValueError("Unrecognized filter name '%s'" % fname)
url = FILTER_URL % fname
data_home = get_data_home(data_home)
if not os.path.exists(data_home):
os.makedirs(data_home)
archive_file = os.path.join(data_home, '%s.dat' % fname)
if not os.path.exists(archive_file):
if not download_if_missing:
raise IOError('data not present on disk. '
'set download_if_missing=True to download')
print "downloading from %s" % url
F = urllib2.urlopen(url)
open(archive_file, 'w').write(F.read())
F = open(archive_file)
return np.loadtxt(F, unpack=True)
def fetch_sdss_filter(fname, data_home=None, download_if_missing=True):
"""Loader for SDSS Filter profiles
Parameters
----------
fname : str
filter name: must be one of 'ugriz'
data_home : optional, default=None
Specify another download and cache folder for the datasets. By default
all scikit learn data is stored in '~/astroML_data' subfolders.
download_if_missing : optional, default=True
If False, raise a IOError if the data is not locally available
instead of trying to download the data from the source site.
Returns
-------
data : ndarray
data is an array of shape (5, Nlam)
first row: wavelength in angstroms
second row: sensitivity to point source, airmass 1.3
third row: sensitivity to extended source, airmass 1.3
fourth row: sensitivity to extended source, airmass 0.0
fifth row: assumed atmospheric extinction, airmass 1.0
"""
if fname not in 'ugriz':
raise ValueError("Unrecognized filter name '%s'" % fname)
url = FILTER_URL % fname
data_home = get_data_home(data_home)
if not os.path.exists(data_home):
os.makedirs(data_home)
archive_file = os.path.join(data_home, '%s.dat' % fname)
if not os.path.exists(archive_file):
if not download_if_missing:
raise IOError('data not present on disk. '
'set download_if_missing=True to download')
print("downloading from %s" % url)
F = urlopen(url)
open(archive_file, 'wb').write(F.read())
F = open(archive_file)
return np.loadtxt(F, unpack=True)
def fetch_rrlyrae_mags(data_home=None, download_if_missing=True):
"""Loader for RR-Lyrae data
Parameters
----------
data_home : optional, default=None
Specify another download and cache folder for the datasets. By default
all scikit learn data is stored in '~/astroML_data' subfolders.
download_if_missing : optional, default=True
If False, raise a IOError if the data is not locally available
instead of trying to download the data from the source site.
Returns
-------
data : recarray, shape = (483,)
record array containing imaging data
Examples
--------
>>> from astroML.datasets import fetch_rrlyrae_mags
>>> data = fetch_rrlyrae_mags() # doctest: +IGNORE_OUTPUT
>>> data.shape # number of objects in dataset
(483,)
Notes
-----
This data is from table 1 of Sesar et al 2010 ApJ 708:717
"""
# fits is an optional dependency: don't import globally
from astropy.io import fits
data_home = get_data_home(data_home)
archive_file = os.path.join(data_home, os.path.basename(DATA_URL))
if not os.path.exists(archive_file):
if not download_if_missing:
raise IOError('data not present on disk. '
'set download_if_missing=True to download')
fitsdata = download_with_progress_bar(DATA_URL)
open(archive_file, 'wb').write(fitsdata)
hdulist = fits.open(archive_file)
return np.asarray(hdulist[1].data)
def fetch_vega_spectrum(data_home=None, download_if_missing=True):
"""Loader for Vega reference spectrum
Parameters
----------
fname : str
filter name: must be one of 'ugriz'
data_home : optional, default=None
Specify another download and cache folder for the datasets. By default
all scikit learn data is stored in '~/astroML_data' subfolders.
download_if_missing : optional, default=True
If False, raise a IOError if the data is not locally available
instead of trying to download the data from the source site.
Returns
-------
data : ndarray
data[0] is the array of wavelength in angstroms
data[1] is the array of fluxes in Jy (F_nu, not F_lambda)
"""
data_home = get_data_home(data_home)
if not os.path.exists(data_home):
os.makedirs(data_home)
archive_name = os.path.join(data_home, VEGA_URL.split('/')[-1])
if not os.path.exists(archive_name):
if not download_if_missing:
raise IOError('data not present on disk. '
'set download_if_missing=True to download')
print "downnloading from %s" % VEGA_URL
F = urllib2.urlopen(VEGA_URL)
open(archive_name, 'w').write(F.read())
F = open(archive_name)
return np.loadtxt(F, unpack=True)
def fetch_vega_spectrum(data_home=None, download_if_missing=True):
"""Loader for Vega reference spectrum
Parameters
----------
fname : str
filter name: must be one of 'ugriz'
data_home : optional, default=None
Specify another download and cache folder for the datasets. By default
all scikit learn data is stored in '~/astroML_data' subfolders.
download_if_missing : optional, default=True
If False, raise a IOError if the data is not locally available
instead of trying to download the data from the source site.
Returns
-------
data : ndarray
data[0] is the array of wavelength in angstroms
data[1] is the array of fluxes in Jy (F_nu, not F_lambda)
"""
data_home = get_data_home(data_home)
if not os.path.exists(data_home):
os.makedirs(data_home)
archive_name = os.path.join(data_home, VEGA_URL.split('/')[-1])
if not os.path.exists(archive_name):
if not download_if_missing:
raise IOError('data not present on disk. '
'set download_if_missing=True to download')
print("downnloading from %s" % VEGA_URL)
F = urlopen(VEGA_URL)
open(archive_name, 'wb').write(F.read())
F = open(archive_name, 'r')
return np.loadtxt(F, unpack=True)
def fetch_sdss_S82standards(data_home=None,
download_if_missing=True,
crossmatch_2mass=False):
"""Loader for SDSS stripe82 standard star catalog
Parameters
----------
data_home : optional, default=None
Specify another download and cache folder for the datasets. By default
all scikit learn data is stored in '~/astroML_data' subfolders.
download_if_missing : bool, optional, default=True
If False, raise a IOError if the data is not locally available
instead of trying to download the data from the source site.
crossmatch_2mass: bool, optional, default=False
If True, return the standard star catalog cross-matched with 2mass
magnitudes
Returns
-------
data : ndarray, shape = (313859,)
record array containing sdss standard stars (see notes below)
Notes
-----
Information on the data can be found at
http://www.astro.washington.edu/users/ivezic/sdss/catalogs/stripe82.html
Data is described in Ivezic et al. 2007 (Astronomical Journal, 134, 973).
Columns are as follows:
RA Right-ascention of source (degrees)
DEC Declination of source (degrees)
RArms rms of right-ascention (arcsec)
DECrms rms of declination (arcsec)
Ntot total number of epochs
A_r SFD ISM extinction (mags)
for each band in (u g r i z):
Nobs_<band> number of observations in this band
mmed_<band> median magnitude in this band
mmu_<band> mean magnitude in this band
msig_<band> standard error on the mean
(1.25 times larger for median)
mrms_<band> root-mean-square scatter
mchi2_<band> chi2 per degree of freedom for mean magnitude
For 2-MASS, the following columns are added:
ra2MASS 2-mass right-ascention
dec2MASS 2-mass declination
J J-band magnitude
Jerr J-band error
H H-band magnitude
Herr H-band error
K K-band magnitude
Kerr K-band error
theta difference between SDSS and 2MASS position (arcsec)
Examples
--------
>>> data = fetch_sdss_S82standards() # doctest: +IGNORE_OUTPUT
>>> u_g = data['mmed_u'] - data['mmed_g']
>>> print(u_g[:4])
[-22.23500061 1.34900093 1.43799973 2.08200073]
References
----------
Ivesic et al. ApJ 134:973 (2007)
"""
data_home = get_data_home(data_home)
if crossmatch_2mass:
archive_file = os.path.join(data_home, ARCHIVE_FILE_2MASS)
data_url = DATA_URL_2MASS
kwargs = dict(dtype=DTYPE_2MASS)
else:
archive_file = os.path.join(data_home, ARCHIVE_FILE)
data_url = DATA_URL
kwargs = dict(usecols=COLUMNS, dtype=DTYPE)
if not os.path.exists(archive_file):
if not download_if_missing:
raise IOError('data not present on disk. '
'set download_if_missing=True to download')
print("downloading cross-matched SDSS/2MASS dataset from %s to %s" %
(data_url, data_home))
zipped_buf = download_with_progress_bar(data_url, return_buffer=True)
gzf = GzipFile(fileobj=zipped_buf, mode='rb')
print("uncompressing file...")
extracted_buf = BytesIO(gzf.read())
data = np.loadtxt(extracted_buf, **kwargs)
np.save(archive_file, data)
else:
data = np.load(archive_file)
return data
def fetch_sdss_sspp(data_home=None, download_if_missing=True, cleaned=False):
"""Loader for SDSS SEGUE Stellar Parameter Pipeline data
Parameters
----------
data_home : optional, default=None
Specify another download and cache folder for the datasets. By default
all scikit learn data is stored in '~/astroML_data' subfolders.
download_if_missing : bool (optional) default=True
If False, raise a IOError if the data is not locally available
instead of trying to download the data from the source site.
cleaned : bool (optional) default=False
if True, then return a cleaned catalog where objects with extreme
values are removed.
Returns
-------
data : recarray, shape = (327260,)
record array containing pipeline parameters
Notes
-----
Here are the comments from the fits file header:
Imaging data and spectrum identifiers for a sample of 327,260
stars with SDSS spectra, selected as:
1) available SSPP parameters in SDSS Data Release 9
(SSPP rerun 122, file from Y.S. Lee)
2) 14 < r < 21 (psf magnitudes, uncorrected for ISM extinction)
3) 10 < u < 25 & 10 < z < 25 (same as above)
4) errors in ugriz well measured (>0) and <10
5) 0 < u-g < 3 (all color cuts based on psf mags, dereddened)
6) -0.5 < g-r < 1.5 & -0.5 < r-i < 1.0 & -0.5 < i-z < 1.0
7) -200 < pmL < 200 & -200 < pmB < 200 (proper motion in mas/yr)
8) pmErr < 10 mas/yr (proper motion error)
9) 1 < log(g) < 5
10) TeffErr < 300 K
Teff and TeffErr are given in Kelvin, radVel and radVelErr in km/s.
(ZI, Feb 2012, [email protected])
Examples
--------
>>> from astroML.datasets import fetch_sdss_sspp
>>> data = fetch_sdss_sspp() # doctest: +IGNORE_OUTPUT
>>> data.shape # number of objects in dataset
(327260,)
>>> print(data.dtype.names[:5]) # names of the first five columns
('ra', 'dec', 'Ar', 'upsf', 'uErr')
>>> print(data['ra'][:1]) # first RA value
[49.6275024]
>>> print(data['dec'][:1]) # first DEC value
[-1.04175591]
"""
# fits is an optional dependency: don't import globally
from astropy.io import fits
data_home = get_data_home(data_home)
archive_file = os.path.join(data_home, os.path.basename(DATA_URL))
if not os.path.exists(archive_file):
if not download_if_missing:
raise IOError('data not present on disk. '
'set download_if_missing=True to download')
fitsdata = download_with_progress_bar(DATA_URL)
open(archive_file, 'wb').write(fitsdata)
hdulist = fits.open(archive_file)
data = np.asarray(hdulist[1].data)
if cleaned:
# -1.1 < FeH < 0.1
data = data[(data['FeH'] > -1.1) & (data['FeH'] < 0.1)]
# -0.03 < alpha/Fe < 0.57
data = data[(data['alphFe'] > -0.03) & (data['alphFe'] < 0.57)]
# 5000 < Teff < 6500
data = data[(data['Teff'] > 5000) & (data['Teff'] < 6500)]
# 3.5 < log(g) < 5
data = data[(data['logg'] > 3.5) & (data['logg'] < 5)]
# 0 < error for FeH < 0.1
data = data[(data['FeHErr'] > 0) & (data['FeHErr'] < 0.1)]
# 0 < error for alpha/Fe < 0.05
data = data[(data['alphFeErr'] > 0) & (data['alphFeErr'] < 0.05)]
# 15 < g mag < 18
data = data[(data['gpsf'] > 15) & (data['gpsf'] < 18)]
# abs(radVel) < 100 km/s
data = data[(abs(data['radVel']) < 100)]
return data