PyFerret 7.3

Release notes are found at http://ferret.pmel.noaa.gov/Ferret/documentation/release-notes/version-7-3-release-notes

Installing pre-built packages

The pre-built distributions are built with HDF5-1.8.18, NetCDF-4.4.1.1, and Cairographics-1.14.10 built in.

The distribution pyferret-7.3-RHEL7-64.tar.gz was build on a RedHat Enterprise Linux v7.3 system. This distribution should be used for 64-bit systems using Python 2.7 (most modern Linux systems) and PyQt4.

The distribution pyferret-7.3-RHEL6-64.tar.gz was build on a RedHat Enterprise Linux v6.9 system. This distribution should be used for 64-bit systems using Python 2.6 (older Linux systems) and PyQt4.

You will need to have the following packages installed using your software manager application, or using a command-line package installation program such as yum or apt-get (which needs to be run as the root user or using the sudo privilege escalation program.)

Required packages may not be already be installed:

• numpy or python-numpy (NumPy)
• libgfortran (Fortran library; if you install SciPy, it will be installed)
• PyQt4 or python-qt4 (Python binding for Qt4; may already be installed)

Highly recommended but optional packages:

• scipy or python-scipy (SciPy)
• pyshp or python-pyshp (PyShp for shapefile functions)

You may also wish to install the netcdf and nco packages to provide some useful programs for working with NetCDF files (such as ncdump and ncattted which are used in the benchmark tests).

If you do not have the Ferret standard datasets, they can be obtained from the NOAA-PMEL/FerretDatasets GitHub repo.

To install, go to the directory to contain the PyFerret subdirectory that this tar file will create and run the command tar xzf pyferret-7.3-RHEL7-64.tar.gz if you downloaded the RHEL7 package, or tar xzf pyferret-7.3-RHEL6-64.tar.gz if you downloaded the RHEL6 package. Then go into the pyferret-7.3-RHEL7-64 or pyferret-7.3-RHEL6-64 subdirectory that was created and run the bin/Finstall script to create the ferret_paths.csh and ferret_paths.sh scripts. (For the PyFerret installation directory, you can answer . - a period representing the current directory.)

To setup the environment variables for running PyFerret, source the appropriate ferret_paths script (the .sh script for bash shells, the .csh script for c-shells). At this point, the pyferret command (script) should successfully start PyFerret and give you the Ferret yes? prompt.

PyFerret 7.2.1 Final Release

(one more rare but important bug fix for v7.2.1)

Installing pre-built packages

The pre-built distributions are built with HDF5-1.8.18, NetCDF-4.4.1.1, and Cairographics-1.14.10 built in.

The distribution pyferret-7.2.1-RHEL7-64.tar.gz was build on a RedHat Enterprise Linux v7.3 system. This distribution should be used for 64-bit systems using Python 2.7 (most modern Linux systems) and PyQt4.

The distribution pyferret-7.2.1-RHEL6-64.tar.gz was build on a RedHat Enterprise Linux v6.9 system. This distribution should be used for 64-bit systems using Python 2.6 (older Linux systems) and PyQt4.

You will need to have the following packages installed using your software manager application, or using a command-line package installation program such as yum or apt-get (which needs to be run as the root user or using the sudo privilege escalation program.)

Required packages may not be already be installed:

• numpy or python-numpy (NumPy)
• libgfortran (Fortran library; if you install SciPy, it will be installed)
• PyQt4 or python-qt4 (Python binding for Qt4; may already be installed)

Highly recommended but optional packages:

• scipy or python-scipy (SciPy)
• pyshp or python-pyshp (PyShp for shapefile functions)

You may also wish to install the netcdf and nco packages to provide some useful programs for working with NetCDF files (such as ncdump and ncattted which are used in the benchmark tests).

If you do not have the Ferret standard datasets, they can be obtained from the NOAA-PMEL/FerretDatasets GitHub repo.

To install, go to the directory to contain the PyFerret subdirectory that this tar file will create and run the command tar xzf pyferret-7.2.1-RHEL7-64.tar.gz if you downloaded the RHEL7 package, or tar xzf pyferret-7.2.1-RHEL6-64.tar.gz if you downloaded the RHEL6 package. Then go into the pyferret-7.2.1-RHEL7-64 or pyferret-7.2.1-RHEL6-64 subdirectory that was created and run the bin/Finstall script to create the ferret_paths.csh and ferret_paths.sh scripts. (For the PyFerret installation directory, you can answer . - a period representing the current directory.)

To setup the environment variables for running PyFerret, source the appropriate ferret_paths script (the .sh script for bash shells, the .csh script for c-shells). At this point, the pyferret command (script) should successfully start PyFerret and give you the Ferret yes? prompt.

Version 7.2.1 Release

PyFerret 7.2.1 Release

This fixes the dynamic memory issues that were discovered in v7.2.0 as well as issues with compiling and running under Mac OS X El Capitan using Homebrew-built libraries. Python memory management is used for all Ferret dynamic memory. Better clean-up is performed by the pyferret.stop method to remove memory leaks and better restore the initial state for a subsequent call to pyferret.start. Apologies for the longer-than-expected delay in getting this update out.

Version 7.2.0 release

PyFerret v7.2.0 release

There is a bug (error) associated with memory used for storing strings (arrays of characters) in this version. This bug may cause the program to crash.

The pre-built distributions are built with HDF5-1.8.18, NetCDF-4.4.1.1, and Cairographics-1.14.10.

The distribution pyferret-7.2.0-RHEL7-64.tar.gz was build on a RedHat Enterprise Linux v7.3 system. This distribution should be used for 64-bit systems using Python 2.7 (most modern Linux systems) and PyQt4.

The distribution pyferret-7.2.0-RHEL6-64.tar.gz was build on a RedHat Enterprise Linux v6.9 system. This distribution should be used for 64-bit systems using Python 2.6 (older Linux systems) and PyQt4.

😞 As usual, there are some updates needed to remove some issues with compiling/running the code on the Mac with homebrew/gcc-7. These actually are generally applicable changes (mostly missing prototypes, one incorrect prototype of no real consequence, calls to malloc/free instead of PyMem_Malloc/PyMem_Free, and a fix to a very rare bug involving an end-of-line wildcard asterisk in the new limited wildcard matching). I hope to get a new release very soon - Karl

Candidate for 7.2 release

Candidate for the v7.2 release of Ferret / PyFerret. Includes dynamic allocation of memory used for variables. So the memory limit set by the user is just a limit on the amount to be allocated and does not cause an immediate allocation of the specified amount of memory.

PyFerret v7.1.1 Initial Release

Beta release of PyFerret

• Using the new memory management
  • each variable allocated memory (using Python memory manager) as needed and when needed
  • SET MEMORY /SIZE=... just specifies a maximum amount of memory to have allocated but does not do any memory allocation itself
• More efficient division of labor when performing computations on large data which can only be partially loaded into memory
• Matrix operations
• Bug fixes

This is being marked as a beta release due to the number of changes to the code, although it does pass all our benchmark tests.

PyFerret v7.1 Final Release

(A few last-minute bugs discovered and fixed since the v7.1 version.)

pyferret-7.1.0-RHEL6-64.tar.gz was built on a RedHat Enterprise Linux 6.8 machine.
This build should be used for 64-bit Linux systems using Python 2.6 and PyQt4.

pyferret-7.1.0-Ubuntu-14.04-64.tar.gz was build on an Ubuntu-14.04 64-bit virtual machine.
This build should be used for 64-bit Linux systems using Python2.7 and PyQt4.

Extract the appropriate tar file into the parent directory of your choice. The tar file will create a subdirectory with a name starting with pyferret-7.1.0. At this time, if you wish, feel free to rename this subdirectory that was created. Then run the bin/Finstall script in this new subdirectory. The directory for FER_DIR should be this new subdirectory. If you do not already have the default Ferret datasets, they can be obtained from NOAA-PMEL/FerretDatasets github repo.

PyFerret v7.1 Release

Release notes:

• Compatible with Python3.6 as well as 2.7 and 2.6
• For displayed plots, uses PyQt5, if available, otherwise PyQt4
• Addition of pyferret.FerFMRCDSet and pyferret.FerFMRCVar objects
• Infinite and NaN values in returned array from the Ferret Python External Function (PyEF) are converted to the missing value for the array

(more still to be written)

Candidate for next PyFerret release

Includes updates found in this version of Ferret
Although not fully tested, now works under both Python2.x and Python3.x and uses either PyQt4 or PyQt5.

Includes the PyFerret object FerFMRCDSet and FerFMRCVar, still to be fully tested.
Still has problems with displayed plots on Mac OS Sierra.

Installing pre-built packages

The PyFerret-7.0.3.2-beta-Ubuntu-14.04-64.tar.gz files should work for most "free" Linux systems using Python2.7 and PyQt4. Unfortunately this is an unoptimized (debug) build, as optimization using updated gcc compilers causes problems for some ferret functionality at this time.

Note that you will need to have the following packages installed using your software manager application, or using the command-line package installation program such as apt-get (which needs to be run using sudo, so the command line will look like sudo apt-get install package-name)

• default-jre (the Java runtime environment, for the ThreddsBrowser) - optional
• python-numpy (NumPy)
• python-scipy (SciPy) - optional, but highly recommended
• python-pyshp (PyShp for shapefile functions) - optional
• libgfortran - if not already installed by the SciPy installation

The following packages were already installed in Ubuntu 14.04, but may not be installed already under other Linux systems.

• python-qt4 (Python binding for Qt4)
• libcurl3 or libcurl4-openssl-dev (for the libcurl.so.4 library)
• libpangocairo (for the pango and pango-cairo text processing libraries)

You may also wish to install the netcdf and nco packages to provide the ncdump and ncattted programs used in the benchmark tests.

If you do not have the Ferret standard datasets, they can be obtained from the github NOAA-PMEL/FerretDatasets repo.

To install, go to the directory to contain the PyFerret subdirectory that this tar file will create and run the command tar xzf PyFerret-7.0.3.2-beta-Ubuntu-14.04-64.tar.gz. Then go into the PyFerret-7.0.3.2-beta subdirectory that was created and run the bin/Finstall script to create the ferret_paths.csh and ferret_paths.sh scripts. (For the PyFerret installation directory, you can answer . - a period representing the current directory.)

To setup the environment variables for running PyFerret, source the appropriate ferret_paths script (the .sh script for bash shells, the .csh script for c-shells). At this point, the pyferret command (script) should successfully start PyFerret.

PyFerret v7.0.0 final release

PyFerret version 7.0.0 release candidate with minor updates.

For the pre-built tar.gz files, you need to extract the appropriate tar.gz file to the directory to contain the PyFerret-7.0.0-... subdirectory. (There is just one tar.gz file for each platform; it contains both the "environment" and "executable" files.)

If you do not already have the default Ferret example datasets installed somewhere, you need to extract the FerretDatasets.tar.gz file to the directory where you want the FerretDatasets directory to reside.

Finally, you need to run the bin/Finstall script, found under the PyFerret-7.0.0-... subdirectory. This script asks for the locations of the PyFerret-7.0.0-... directory and the FerretDatasets directory, and then creates the ferret_paths scripts.

To run PyFerret, you first need to set Ferret environment variable by "source"-ing the appropriate ferret_paths script (". ferret_paths.sh" for Bourne shells, "source ferret_paths.csh" for C-shells). Once these environment variables are set, you should be able to run the pyferret program (script). It should be automatically found (from the PATH environment variables), but resides in the bin subdirectory of your PyFerret-7.0.0-... directory.