1. Python Crash Course PyFits, Astropy
3rd year Bachelors
V1.0 dd
Hour 7

2. Handling FITS files - PyFITS
Read, write and manipulate all aspects of FITS files
extensions
headers
images
tables
Low-level interface for details
High-level functions for quick and easy use

3. PyFITS - reading >>> import pyfits
>>> imgname = “testimage.fits”
>>> img = pyfits.getdata(imgname)
>>> img
array([[2408, 2408, 1863, ..., 3660, 3660, 4749],
[2952, 2408, 1863, ..., 3660, 3115, 4204],
[2748, 2748, 2204, ..., 4000, 3455, 4000],
...,
[2629, 2901, 2357, ..., 2261, 2806, 2261],
[2629, 2901, 3446, ..., 1717, 2261, 1717],
[2425, 2697, 3242, ..., 2942, 2125, 1581]], dtype=int16)
>>> img.mean()
>>> img[img > 2099].mean()
>> import numpy
>>> numpy.median(img)
4244.0

4. PyFITS – reading FITS images
>>> x = 348; y = 97
>>> delta = 5
>>> print img[y-delta:y+delta+1,
... x-delta:x+delta+1].astype(numpy.int)
[[ ]
[ ]
[ ]
[ ]
[ ]
[ ]
[ ]
[ ]
[ ]
[ ]
[ ]]
row = y = first index
column = x = second index
numbering runs as normal (e.g. in ds9) BUT zero indexed!

5. PyFITS – reading FITS tables
>>> tblname = ‘data/N891PNdata.fits’
>>> d = pyfits.getdata(tblname)
>>> d.names
('x0', 'y0', 'rah', 'ram', 'ras', 'decd', 'decm', 'decs', 'wvl', 'vel',
'vhel', 'dvel', 'dvel2', 'xL', 'yL', 'xR', 'yR', 'ID', 'radeg', 'decdeg',
'x', 'y')
>>> d.x0
array([ , , , ,…
, , ], dtype=float32)
>>> d.field(‘x0’) # case-insensitive
>>> select = d.x0 < 200
>>> dsel = d[select] # can select rows all together
>>> print dsel.x0 [ ]

6. PyFITS – reading FITS headers
>>> h = pyfits.getheader(imgname)
>>> print h
SIMPLE = T /FITS header
BITPIX = /No.Bits per pixel
NAXIS = /No.dimensions
NAXIS1 = /Length X axis
NAXIS2 = /Length Y axis
EXTEND = T /
DATE = '05/01/ ' /Date of FITS file creation
ORIGIN = 'CASB -- STScI ' /Origin of FITS image
PLTLABEL= 'E ' /Observatory plate label
PLATEID = '06UL ' /GSSS Plate ID
REGION = 'XE ' /GSSS Region Name
DATE-OBS= '22/12/ ' /UT date of Observation
UT = '03:09: ' /UT time of observation
EPOCH = E+03 /Epoch of plate
PLTRAH = /Plate center RA
PLTRAM = /
PLTRAS = E+00 /
PLTDECSN= ' ' /Plate center Dec
PLTDECD = /
PLTDECM = / >>> h[‘KMAGZP’]
>>> h['REGION']
'XE295‘
# Use h.items() to iterate through all header entries

7. PyFITS – writing FITS images
>>> newimg = sqrt((sky+img)/gain + rd_noise**2) * gain
>>> newimg[(sky+img) < 0.0] = 1e10
>>> hdr = h.copy() # copy header from original image
>>> hdr.add_comment(‘Calculated noise image’)
>>> filename = ‘sigma.fits’
>>> pyfits.writeto(filename, newimg, hdr) # create new file
>>> pyfits.append(imgname, newimg, hdr) # add a new FITS extension
>>> pyfits.update(filename, newimg, hdr, ext) # update a file
# specifying a header is optional,
# if omitted automatically adds minimum header

8. PyFITS – writing FITS tables
>>> import pyfits
>>> import numpy as np
>>> # create data
>>> a1 = numpy.array(['NGC1001', 'NGC1002', 'NGC1003'])
>>> a2 = numpy.array([11.1, 12.3, 15.2])
>>> # make list of pyfits Columns
>>> cols = []
>>> cols.append(pyfits.Column(name='target', format='20A',
array=a1))
>>> cols.append(pyfits.Column(name='V_mag', format='E', array=a2))
>>> # create HDU and write to file
>>> tbhdu=pyfits.new_table(cols)
>>> tbhdu.writeto(’table.fits’)
# these examples are for a simple FITS file containing just one
# table or image but with a couple more steps can create a file
# with any combination of extensions (see the PyFITS manual online)

9. PyFITS - advanced >>> f = pyfits.open(tblname)
>>> f.info()
Filename: data/N891PNdata.fits
No. Name Type Cards Dimensions Format
0 PRIMARY PrimaryHDU () uint8
BinTableHDU R x 22C [E, E, E, E, E,
E, E, E, E, E, E, E, E, E, E, E, E, E, E, E, E, E]
>>> table = f[1] # data extension number 1 (can also use names)
>>> d = f[1].data # data, same as returned by pyfits.getdata()
>>> h = f[1].header # header, same as returned by pyfits.getheader ()
>>> # make any changes
>>> f.writeto(othertblname) # writes (with changes) to a new file
>>> f = pyfits.open(tblname, mode=‘update’) # to change same file
>>> f.flush() # writes changes back to file
>>> f.close() # writes changes and closes file

10. Astropy
The astropy package (alternatively known as the “core” package)
contains various classes, utilities, and a packaging framework
intended to provide commonly-used astronomy tools. It is divided
into a variety of sub-packages (
>>> import astropy
>>> from astropy.io import fits
In [1]: import astropy.
astropy._compiler astropy.io astropy.tests
astropy.builtins astropy.logger astropy.time
astropy.config astropy.logging astropy.units
astropy.conftest astropy.nddata astropy.utils
astropy.constants astropy.setup_helpers astropy.version
astropy.coordinates astropy.sphinx astropy.version_helpers
astropy.cosmology astropy.stats astropy.wcs
astropy.cython_version astropy.sys
astropy.extern astropy.table

11. Astropy Constants
astropy.constants contains a number of physical constants useful in Astronomy. Constants are Quantity objects with additional meta-data describing their provenance and uncertainties.
>>> from astropy.constants import G
>>> from astropy import constants as const
>>> print const.c
Name = Speed of light in vacuum
Value =
Error = 0.0
Units = m / (s)
Reference = CODATA 2010
>>> print const.c.to('pc/yr')
pc / (yr)
>>> F = (const.G * 3. * const.M_sun * 100 * u.kg) / (2.2 * u.au) ** 2
>>> print F.to(u.N) N

12. Astropy Units
astropy.units handles defining and converting between physical units, and performing arithmetic with physical quantities (numbers with associated units).
>>> from astropy import units as u
>>> u.pc.to(u.m)
e+16
>>> cms = u.cm / u.s
>>> mph = u.mile / u.hour
>>> cms.to(mph, 1)
>>> cms.to(mph, [1., 1000., 5000.])
array([ e-02, e+01, e+02])
>>> u.nm.to(u.Hz, [1000, 2000])
UnitsException: 'nm' (length) and 'Hz' (frequency) are not convertible
>>> u.nm.to(u.Hz, [1000, 2000], equivalencies=u.spectral())
array([ e+14, e+14])
>>> F = (const.G * 3. * const.M_sun * 100 * u.kg) / (2.2 * u.au) ** 2
>>> print F.to(u.N) N

13. Astropy nddata
astropy.nddata provides the NDData class and related tools to manage n-dimensional array-based data (e.g. CCD images, IFU data, grid-based simulation data, ...). This is more than just numpy.ndarray objects, because it provides metadata that cannot be easily provided by a single array
>>> from astropy.nddata import NDData
>>> array = np.random.random((12, 12, 12)) # a random 3-dimensional array
>>> ndd = NDData(array)
>>> ndd.ndim 3
>>> ndd.shape
(12, 12, 12)
>>> ndd.dtype
dtype('float64')
>>> ndd.data # the underlying numpy array
array([[[ , , , ..., , , ],
[ , , , ..., ,

14. Astropy nddata
Nndata handles Masks, Flags, Uncertainties and Meta data
>>> ndd.mask = ndd.data > 0.9
>>> ndd.flags = np.zeros(ndd.shape)
>>> ndd.flags[ndd.data < 0.1] = 1
# multiple flag layers
>>> from astropy.nddata import FlagCollection
>>> ndd.flags = FlagCollection(shape=(12, 12, 12))
>>> ndd.flags['photometry'] = np.zeros(ndd.shape, dtype=str)
>>> ndd.flags['photometry'][ndd.data > 0.9] = 's'
>>> ndd.flags['cosmic_rays'] = np.zeros(ndd.shape, dtype=int)
>>> ndd.flags['cosmic_rays'][ndd.data > 0.99] = 99
>>> from astropy.nddata import StdDevUncertainty
>>> ndd.uncertainty = StdDevUncertainty(np.ones((12, 12, 12)) * 0.1)
>>> ndd.meta['exposure_time'] = 340.
>>> ndd.meta['filter'] = 'J‘
# back to numpy array
>>> arr = np.array(ndd)

15. Astropy nddata
astropy.nddata includes a convolution function that offers improvements compared to the scipy astropy.ndimage convolution routines, including:
Proper treatment of NaN values
A single function for 1-D, 2-D, and 3-D convolution
Improved options for the treatment of edges
Both direct and Fast Fourier Transform (FFT) versions
from astropy.nddata import convolve, convolve_fft
result = convolve(image, kernel)
result = convolve_fft(image, kernel)
>>> convolve([1, 4, 5, 6, np.nan, 7, 8],
[0.2, 0.6, 0.2],
boundary='extend')
array([ 1.6, 3.6, 5. , 5.9, 6.5, 7.1, 7.8])

16. Astropy tables
Astropy.table provides functionality for storing and manipulating heterogenous tables of data in a way that is familiar to numpy users. A few notable features of this package are:
Initialize a table from a wide variety of input data structures and types.
Modify a table by adding or removing columns, changing column names, or adding new rows of data.
Handle tables containing missing values.
Include table and column metadata as flexible data structures.
Specify a description, units and output formatting for columns.
Interactively scroll through long tables similar to using more.
Create a new table by selecting rows or columns from a table.
Full support for multidimensional columns.
Methods for Reading and writing Table objects to files

17. Astropy tables >>> from astropy.table import Table, Column
>>> c = ['x', 'y', 'z']
>>> t = Table([a, b, c], names=('a', 'b', 'c'), meta={'name': 'first table'})
>>> print t
a b c x y z
>>> t['a'] # Column 'a'
<Column name='a' units=None format=None description=None>
array([1, 4, 5])
>>> t['a'][1] # Row 1 of column 'a' 4
>>> t.add_column(Column(data=[1, 2, 3], name='d')))
>>> t.remove_column('c')

18. Astropy tables >>> t = Table()
>>> t.add_column(Column(data=[1, 4], name='a'))
>>> t.add_row((1, 2.0, ‘b'))
>>> Table([t['a']**2, t['b'] + 10])
<Table rows=2 names=('a','b')>
array([(1, 16.0), (11, 12.0)],
dtype=[('a', '<i8'), ('b', '<f8')])
>>> arr = {'a': [1, 4],
'b': [2.0, 5.0],
'c': ['x', 'y']}
>>>
>>> Table(arr)
<Table rows=2 names=('a','c','b')>
array([(1, 'x', 2.0), (4, 'y', 5.0)],
dtype=[('a', '<i8'), ('c', '|S1'), ('b', '<f8')])
>>> t['b'].mask = [True, False] # Modify column mask (boolean array)
>>> print(t)
a b
1 --
-- 4

19. Astropy time
The astropy.time package provides functionality for manipulating times and dates. Specific emphasis is placed on supporting time scales (e.g. UTC, TAI, UT1) and time representations (e.g. JD, MJD, ISO 8601) that are used in astronomy. It uses Cython to wrap the C language SOFA (standards of Fundamental astronomy, IAU) time and calendar routines.
Scale
Description
tai
International Atomic Time (TAI)
tcb
Barycentric Coordinate Time (TCB)
tcg
Geocentric Coordinate Time (TCG)
tdb
Barycentric Dynamical Time (TDB) tt
Terrestrial Time (TT)
ut1
Universal Time (UT1)
utc
Coordinated Universal Time (UTC)
>>> from astropy.time import Time
>>> times = [' :00: ', ' :00:00']
>>> t = Time(times, format='iso', scale='utc')
>>> t
<Time object: scale='utc' format='iso' vals=[' :00:00.123' ' :00:00.000']>
>>> t.jd
array([ , ])
>>> t.mjd
array([ , ]
>>> t1 = Time(' :00:00', scale='utc')
>>> t2 = Time(' :00:00', scale='utc')
>>> dt = t2 - t1 # Difference between two Times
>>> dt
<TimeDelta object: scale='tai' format='jd' vals=31.0>
>>> dt.sec

20. Astropy io
Astropy provides a unified interface for reading and writing data in different formats. For many common cases this will simplify the process of file I/O and reduce the need to master the separate details of all the I/O packages within Astropy.
>>> from astropy.table import Table
>>> t = Table.read('2mass.tbl', format='ipac')
>>> t = Table.read('aj285677t3.txt', format='cds')
>>> t = Table.read('photometry.dat', format='daophot')
>>> t = Table.read('paper_table.tex')
>>> t = Table.read('photometry.dat', format='ascii', data_start=2, delimiter='|')
>>> t = Table.read('observations.hdf5', path='group/data')
>>> t = Table.read('catalog.xml', table_id='twomass', format='votable')

21. Astropy io
The astropy.io.fits package provides access to FITS files. FITS (Flexible Image Transport System) is a portable file standard widely used in the astronomy community to store images and tables.
>>> from astropy.io import fits
>>> hdulist = fits.open('input.fits')
>>> hdulist.close()
>>> hdulist.info()
Filename: test1.fits
No. Name Type Cards Dimensions Format
0 PRIMARY PrimaryHDU () int16
1 SCI ImageHDU (800, 800) float32
2 SCI ImageHDU (800, 800) float32
>>> prihdr = hdulist[0].header
>>> prihdr['targname'] = ('NGC121-a', 'the observation target')
>>> prihdr['targname']
'NGC121-a'
>>> prihdr.comments['targname']
'the observation target'

22. Astropy io
Image data
>>> scidata = hdulist[1].data # pointer to the data
>>> scidata.shape
(800, 800)
>>> scidata.dtype.name
'float32‘
>>> scidata[30:40, 10:20]
# array manipulation to convert the image data from counts to flux:
>>> photflam = hdulist[1].header['photflam']
>>> exptime = prihdr['exptime']
>>> scidata *= photflam / exptime
>>> hdulist.writeto('newimage.fits')
>>> f = fits.open('original.fits', mode='update')
... # making changes in data and/or header
>>> f.flush() # changes are written back to original.fits

23. Astropy WCS
framework to represent celestial coordinates and transform between them
>>> from astropy import coordinates as coord
>>> from astropy import units as u
>>> coord.ICRSCoordinates(ra= , dec= , unit=(u.degree, u.degree))
<ICRSCoordinates RA= deg, Dec= deg>
>>> coord.ICRSCoordinates('00h42m44.3s +41d16m9s')
>>> c = coord.ICRSCoordinates(ra= , dec= , unit=(u.degree, u.degree))
>>> c.ra
<RA deg>
>>> c.ra.hours
>>> c.ra.hms
(0.0, 42, )
>>> c.dec
<Dec deg>
>>> c.dec.radians

24. Astropy WCS Coordinate transformations >>> c.galactic
<GalacticCoordinates l= deg, b= deg>
>>> c.transform_to(coord.GalacticCoordinates)
# Lookup object names at CDS, requires internet connection
>>> c_eq = coord.ICRSCoordinates.from_name("M33")
>>> c_eq
<ICRSCoordinates RA= deg, Dec= deg>

25. Introduction to language
End