WaveThresh4 Beta Software

The following software is LEGACY and only provided for historical use. The software won't work properly with R2.9.0 and later. Use the newer versions on the WaveThresh home page The installation advice might be useful for the recent packages as well.

oDownload WaveThresh4 BETA for R for Windows (Zip file) (v4.0.2)
oDownload WaveThresh4 BETA for R for Linux (Gzipped tarball) (v4.0.3)
oDownload WaveThresh4 BETA for R for Mac OS X Intel (Gzipped tarball) (v4.0.3)
o(Download WaveThresh4 BETA for R for Mac OS X PowerPC (Gzipped tarball) (v4.0.2))
Installation is easy, see instructions below.

Since this is a beta version we would appreciate knowing about any bugs or improvements. Any problems please contact Dan Bailey

Installation Instructions

On Windows

  1. Download the file pointed to by the above link (for Windows). Save the file wavethresh4_4.0-2.zip to a convenient location.
  2. Start R.
  3. Using the menu items select Packages and then select Install packages(s) from local zip files and then choose the zip file that you downloaded. Wait until it finishes.
  4. Then whenever you wish to use the package you need to type library("wavethresh")
  5. After ending your R session if you wish to restart R and use the package again then you need to repeat step 4. every time you wish to use WaveThresh.

On Linux

The Linux version should work for both 32 and 64 bit systems.

On Linux you need to decide whether this is going to be a system-wide installation (in which case you need the appropriate permission) or a personal installation (ie, the package files get stuffed somewhere under your home directory).

Personal installation

  1. First, decide on where you want personal packages to be stored and create that subdirectory. For example, I use the directory /home/guy/myrlibrary you should replace this string by whatever you decide to use (and make sure the directory exists).
  2. If you haven't already, create a file called .Rprofile in your home directory. In that file add the line
    remembering to replace the directory by whatever you chose above.
  3. Download and save the Gzipped tarball wavethresh4_4.0-2.tar.gz to a convenient location.
  4. In the same directory where the tarball is located type
    R CMD INSTALL wavethresh4_4.0-2.tar.gz
    and wait whilst the package is installed. (It gets installed by whatever directory you put in the .libPaths command in the .Rprofile file.
  5. After installation to use WaveThresh you need to start R, and then type library("wavethresh") every time you wish to use it.
Systemwide installation I have not done this personally however, I think you can do this by making sure that you have the appropriate permissions (e.g. you might have to be the root user) and then follow steps 3 and 4 from the Personal Installation instructions above. You need to make sure that there is no .libPaths command in any .Rprofile file which diverts the install to another directory.

On the Intel Mac

You should just be able to save the file somewhere and then type
R CMD INSTALL wavethresh4_4.0-3.tar.gz
See also instructions for Linux above.

On the PowerPC Mac

You should just be able to save the file somewhere and then type
R CMD INSTALL wavethresh4_4.0-2.tar.gz
although note that the installation does require the g77 FORTRAN compiler. You may already have this installed. If you do not then it is trivial to install g77 using the Fink. I use the Fink Commander package as an easy-to-use GUI for Fink.

cc_dynamic problems
One user experienced problems with installation and got the error message
can't locate file for: -lcc_dynamic
His solution was to edit the Makefile and remove the -lcc_dynamic flag and then remake (the *.so) and that seemed to work.

Apparently the problem is to do with different versions of gcc being used. A web newsgroup suggested to run sudo gcc_select 3.3 which changes the version of the default gcc to be 3.3. Although from reading other material on the web it seems like this should not be an issue when using versions of R>2.2.0. The quote is: R 2.2.0 CRAN release provides a fall-back soft link to cc_dynamic such that gcc_select should not be necessary anymore. (from this source)