[AstroPy] New Version of STScI Science Software Available
Warren Hack
hack at stsci.edu
Fri Apr 29 00:05:16 EDT 2016
********************************************************
SSB Science Software NOW AVAILABLE
********************************************************
The Science Software Branch of the Space Telescope Science Institute
wishes to announce the availability of AstroConda, a replacement for the
previously released Ureka environment along with updates to many of the packages
installed as part of the STScI_Python releases. Full details on the new software environment
and how to install it can be found at:
http://astroconda.readthedocs.io
-----------
SSBX
-----------
The transition to AstroConda had one unexpected consequence that only appeared
the week of April 25; specifically, the Ureka-based SSBX environment can no
longer be installed. We have two options available for those who have relied on
the SSBX environment:
* Start using the (less stable) Ureka SSBDEV environment
- Instructions on how to install and use this environment can be found at:
http://ssb.stsci.edu/ssb_software.shtml
* Transition to using AstroConda, our Conda-based Python environment that
replaces the Ureka environment
- The equivalent of SSBX can be maintained by following the instructions
for "Updating All Packages" as described on:
http://astroconda.readthedocs.io/en/latest/updating.html
We regret the problems this unexpectedly forced transition has caused for many
in the STScI community, but all the lessons learned from providing and
maintaining Ureka over the last few years have been applied to our new
distribution. AstroConda is more stable, and will not disappear like
Ureka's SSBX environment did; as problems with any package will simply result in
a lack of updates for that package until the problems are resolved by the
developer.
-----------
IRAF
-----------
We continue to provide IRAF as a component of this new AstroConda environment.
However, more modern computer environments have placed increasing pressure on our
ability to build IRAF for new operating systems. The current installation includes
IRAF v2.16.1 built under OS X 10.6 (for the Mac install) and under
Red Hat Enterprise Linux 5 (for the Linux install). Instructions are provided to
allow you to install IRAF under the Python 2.7 environment:
http://astroconda.readthedocs.io/en/latest/faq.html
These binaries, however, are known to be incompatible with some newer systems
based on reports from users; specific configurations that produce these problems are not
clearly identified, but (as one example only) a Linux system with an XFS disk partition
larger than 2Tb has been seen to be incompatible with IRAF. Should you find
this version of IRAF (should you choose to install it) does not work, we have
developed a Virtual Machine install for all the IRAF-related software (contact us
for details should you need this solution).
We plan to enhance the IRAF environment so it may be installed in a more modular fashion
using Conda (under AstroConda). This enables Gemini and other contributors relying
on IRAF to provide updates without requiring re-installation of the entire IRAF
environment. Unfortunately, the non-VM/native installation of IRAF can only be
supported for as long as we are able to build and run IRAF, which we anticipate
will be possible for at least the next year or two barring any unforeseen OS/compiler
changes.
-----------
Python Support
-----------
The AstroConda environment provides installations for Python 2.7, Python 3.4,
and Python 3.5. Unfortunately, at this time, the IRAF installation can only be
supported under Python 2.7.
-----------
Installation
-----------
Full installation instructions, and explanations, can be found on the AstroConda
web page. However, should you wish to jump head-long into the new Conda-based
environment, you can follow these simple steps:
1. Install Anaconda locally on your system
2. Start a BASH shell - the entire conda environment inextricably relies on BASH
3. Point to our AstroConda channel using:
conda config --add channels http://ssb.stsci.edu/astroconda
3. Install the software from the AstroConda channel using:
conda create -n astroconda stsci
4. Start it up (under the BASH command shell) with:
source activate astroconda
Please go the web page for full details on how to work best with this environment.
-----------
Support
-----------
Questions about this installation and the STScI-based and Gemini-based software
included in this environment can be directed to:
* EMAIL: help at stsci.edu
* PHONE: 410-338-4400
* Web:
# AstroConda:
http://astroconda.readthedocs.io
# For stsci_python documentation:
http://ssb.stsci.edu/doc/stsci_python_dev/
================
Updated Software
================
The full set of release notes for all the STScI software included in this
installation can be found at:
http://astroconda.readthedocs.io/en/latest/release_notes.html
These notes provide expanded details for updates and bug fixes implemented for the
following major packages. Please see the web page for the full set of release notes.
-----------
Dependencies and SciPy
-----------
An effort has been made to start replacing internal dependencies on stsci.*
packages where equivalents are available through SciPy. One example would be
stsci.convolve, which was an early fork of scipy's convolve package, with the
stsci_python code being modified to use scipy instead. As a result, we will
start deprecating some of these packages in future releases for eventual removal
from our distribution. After all, who needs our version when a better
scipy version exists? The names of those deprecated packages will be clearly
noted when we start this transition, and we will not remove these packages for
at least 6 months after we announce their deprecation.
-----------
CALWF3
-----------
CALWF3 Version 3.3 applies instrument calibrations to HST/WFC3 data. This version
includes the following updates and bug fixes (and more...):
* Addition of a new processing step, FLUXCORR, to the UVIS pipeline
* Addition of CTE correction for full-frame UVIS data
* sink pixel detection is now performed in the UVIS pipeline for full-frame UVIS images
* CRCORR processing turned off by default for IR SCAN data
* assorted bug fixes (including memory leaks)
-----------
CALACS
-----------
CALACS Version v8.3.3 applies instrument calibrations to HST/ACS data. This version
includes the following updates and bug fixes (and more):
* Added support for 2K subarrays in PCTECORR
* Improved temporary file handling.
* Added support for very long input list for ACSREJ.
-----------
CALCOS
-----------
CALCOS Version 3.1.3 applies instrument calibrations to HST/COS data. This version
includes the following updates:
* Revisions needed to allow calcos to run under Python 3
* Allow flux extraction from 0 to 100%
* Ignore DQ flag DQ_GAIN_SAG_HOLE in background regions in profile alignment step
* Shift DQ data array the same as data array in TRCECORR and ALGNCORR steps of calcos
* Fix up x1d output for boxcar option
* CalCOS should print warnings for certain conditions of HOTSPOT calibration
* Events in COS data should now be flagged using the SPOTTAB
-----------
DrizzlePac
-----------
The DrizzlePac package, a component of the STScI_Python environment,
implements the distortion correction and image combination algorithm for
HST data. This release of DrizzlePac Version 2 updates the software to provide
a number of new capabilities and bug fixes; including,
* support for automated mosaic building
* improved sky matching
* support for the simplified/improved ACS time-dependent distortion correction
* allowing undistorted non-HST images to be used as references for aligning HST images
* better capability for aligning images from different HST cameras
* ability to run under Python 2.7 and Python 3.5
A full set of release notes can be found online at the DrizzlePac home page at:
http://drizzlepac.stsci.edu
-----------
PySynphot
-----------
This release includes pysynphot v0.9.8.2. The following changes (and a
little more) have been made since the last release of v0.9.7 (Oct 2015):
* Updated spectra data including ACS wavecat and Vega reference spectrum.
* Replaced PyFITS dependency with astropy.io.fits.
* Added a lot of documentation and tutorials
More information about the AstroPy
mailing list