The SAS SWAT package is a Python interface to the SAS Cloud Analytic Services (CAS) engine (the centerpiece of the SAS Viya framework). With this package, you can load and analyze data sets of any size on your desktop or in the cloud. Since CAS can be used on a local desktop or in a hosted cloud environment, you can analyze extremely large data sets using as much processing power as you need, while still retaining the ease-of-use of Python on the client side.
Using SWAT, you can execute workflows of CAS analytic actions, then pull down the summarized data to further process on the client side in Python, or to merge with data from other sources using familiar Pandas data structures. In fact, the SWAT package mimics much of the API of the Pandas package so that using CAS should feel familiar to current Pandas users.
With the best-of-breed SAS analytics in the cloud and the use of Python and its large collection of open source packages, the SWAT package gives you access to the best of both worlds.
To access the CAS binary protocol (recommended), you need the following:
- 64-bit Python 2.7.x or 3.4+ on Linux (see shared library notes below)
The binary protocol requires pre-compiled components found in the
pip installer only. These pieces are not available as source code and
are under a separate license (see documentation on SAS TK). The binary protocol
offers better performance than REST, especially when transferring larger
amounts of data. It also offers more advanced data loading from the client
and data formatting features.
To access the CAS REST interface only, you can use the pure Python code which runs in Python 2.7/3.4+ on all platforms. While not as fast as the binary protocol, the pure Python interface is more portable.
If you do not have pip installed, you can use easy_install pip to add
it to your current Python installation.
Some Linux distributions may not install all of the needed shared libraries
by default. Most notably, the shared library libnuma.so.1 is required to
make binary protocol connections to CAS. If you do not have this library on
your machine you can install the numactl package for your distribution
to make it available to SWAT.
The SWAT package uses many features of the Pandas Python package and other
dependencies of Pandas. If you do not already have version 0.16.0 or greater
of Pandas installed, pip will install or update it for you when you
install SWAT.
SWAT can be installed from the
SWAT project releases page.
Simply locate the file for your platform and install it using pip as
follows:
pip install https://github.com/sassoftware/python-swat/releases/download/vX.X.X/python-swat-X.X.X-platform.tar.gz
Where X.X.X is the release you want to install, and platform is the
platform you are installing on. You can also use the source code distribution
if you only want to use the CAS REST interface. It does not contain support
for the binary protocol.
For the full documentation go to sassoftware.github.io/python-swat. A simple example is shown below.
Once you have SWAT installed and you have a CAS server to connect to, you can import swat and create a connection::
>>> import swat
>>> conn = swat.CAS(host, port, username, password)
If that is successful, you should be able to run an action on the CAS server::
>>> out = conn.serverstatus()
NOTE: Grid node action status report: 1 nodes, 6 total actions executed.
>>> print(out)
[About]
{'CAS': 'Cloud Analytic Services',
'Copyright': 'Copyright © 2014-2016 SAS Institute Inc. All Rights Reserved.',
'System': {'Hostname': 'cas01',
'Model Number': 'x86_64',
'OS Family': 'LIN X64',
'OS Name': 'Linux',
'OS Release': '2.6.32-504.12.2.el6.x86_64',
'OS Version': '#1 SMP Sun Feb 1 12:14:02 EST 2015'},
'Version': '3.01',
'VersionLong': 'V.03.01M0D08232016',
'license': {'expires': '20Oct2016:00:00:00',
'gracePeriod': 62,
'site': 'SAS Institute Inc.',
'siteNum': 1,
'warningPeriod': 31}}
[server]
Server Status
nodes actions
0 1 6
[nodestatus]
Node Status
name role uptime running stalled
0 cas01 controller 4.836 0 0
+ Elapsed: 0.0168s, user: 0.016s, sys: 0.001s, mem: 0.287mb
>>> conn.close()
Copyright SAS Institute