The FastCSWT package provides functionality to perform a directional continuous wavelet transform on the sphere. The transform is based on the construction of the continuous spherical wavelet transform (CSWT) developed by Antoine and Vanderghynest 1999. A fast implementation of the CSWT (based on the fast spherical convolution developed by Wandelt and Gorski 2001) is also provided. The fast algorithm in the context of the CSWT, and as implemented in this package, is described in our paper:
J. D. McEwen, M. P. Hobson, D. J. Mortlock, and A. N. Lasenby. Fast directional continuous spherical wavelet transform algorithms. IEEE Trans. Sig. Proc., 55(2):520-529, 2007.
The library design is based on object-oriented principles and is implemented in Fortran 90. Generally one must instantiate a class object using the corresponding init routines. The instatiated object should be freed once it is no longer required with the corresponding free routine (this deallocates all memory associated with the object).
For further details see the README.txt file.
The FastCSWT package requires the S2, HEALPix, CFITSIO and FFTW 2.1.5 packages. Please note that all FastCSWT code is written in Fortran 90 and a suitable Fortran compiler will be required.
>> make lib
For details on how to use the FastCSWT library code see the documentation
below.
The FastCSWT utility programs may be compiled by running:
By default, the FastCSWT code ships with only this single documentation
page. The remainder of the documentation is automatically generated
from the source code using
f90doc.
Please ensure that you have f90doc installed in order to generate this
documentation. Once f90doc is installed the documentation may be
generated by running:
Compiling utility programs
For details on how to use the FastCSWT utility programs see the documentation
below.
>> make prog
Documentation
>> make docs
Cleaning up
To clean up your version of the FastCSWT code and return all code to its
original state run:
To remove all documentation, except the front page (i.e. this file),
run:
>> make clean
>> make cleandocs
Library classes
An overview of the classes that comprise the FastCSWT library is given
here. Please click on the link for each class for detailed
documentation.
cswt_tr_mod
:
Provides functionality to support and compute the directional
spherical wavelet transform. In addition the mother wavelet, the
dilation values and all wavelet parameters are stored herein. The
resulting wavelet coefficients for each dilation are sampled on an
ecp (equi-sampled) alpha-beta-gamma Euler angle grid. Functionality
is also provided to convert the alpha-beta dimensions to Healpix sky
representations and vice versa.
cswt_swav_mod
:
Provides functionality to support and manipulate a spherical
wavelet. The spherical wavelet data structure includes the wavelet
sky plus its parameters.
cswt_tmpl_mod
:
Contains definitions of template functions defined on both the sky
and the plane used to initialise spherical wavelets. The functions
defined on the plane are projected onto the sphere numerically,
whereas those functions defined on the sphere have already been
analytically projected from the plane.
cswt_error_mod
:
Functionality to handle errors that may occur in the FastCSWT
library. Public FastCSWT error codes are defined, with corresponding
private error comments and default halt execution status.
cswt_analysis
:
Perform fast directional continuous spherical wavelet transform.
cswt_mask_apply
:
Apply an extended coefficient exclusion mask to wavelet coefficients.
cswt_mask_copy
:
Generate a coefficient exclusion mask defined in the ecp wavelet
coefficient domain by simply converting a list of Healpix sky mask to
the wavelet domain.
cswt_mask_gen
:
Generate an extended coefficient exclusion mask defined in the ecp
wavelet coefficient domain from an original mask defined on the sky
and the spherical wavelet transform of the original mask.
cswt_mask_invert
:
Invert a coefficient mask by converting ones to zeros, and zeros to
ones.
cswt_mask_nonzero
:
Compute number of effective coefficients in an extended coefficient
exclusion mask and write results to either the standard output or to
a file.
cswt_mask_nonzero_weight
:
Compute the weighted number of effecitve coefficients in the extended
coefficient mask. Weights for each pixel are given by the size of the
pixel on the sky for the ECP wavelet coefficient domain.
cswt_plot_swav
:
Plot spherical wavelet for a range of dilations.
cswt_swav_azbandlim
:
Determine azimuthal band limit of wavelet.
cswt_tr2sky
:
Convert a spherical wavelet transform fits file (ecp sampled) to a
sky Healpix fits file. Just the sky corresponding to a specified
dilation and orientation index may be written, or skies
corresponding to all dilations and orientations may be written.
cswt_tr_nsigma
:
Convert wavelet coefficients to nsigma values. The coefficient map
for each dilation is considered; the mean for each dilation is
subtracted, before dividing by the sigma value computed for that
dilation. Note that the mean and sigma are only computed over the
non-masked pixels.
J. D. McEwen, M. P. Hobson, D. J. Mortlock, and A. N. Lasenby. Fast directional continuous spherical wavelet transform algorithms. IEEE Trans. Sig. Proc., 55(2):520-529, 2007.
Author: J. D. McEwen
Version: 1.0 - July 2008
FastCSWT package to provide functionality to perform a directional continuous wavelet transform on the sphere
Copyright (C) 2005 Jason McEwenThis program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details (LICENSE.txt).
You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA.
JDM would like to thank Daniel Morlet for providing the code base for the implementation of Wandelt and Gorski's fast spherical convolution, as well as for useful discussions.
This documentation has been generated largely by f90doc, with some minor modifications.
Last modified: 1 July 2008