A Python client to remotely access the R statistic package via network
What It Does
pyRerve is a library for connecting Python to an R process (an excellent statistic package) running Rserve as a RPC connection gateway. Through such a connection variables can be get and set in R from Python, and also R-functions can be called remotely. In contrast to rpy or rpy2 the R process does not have to run on the same machine, it can run on a remote machine and all variable access and function calls will be delegated there through the network.
Furthermore - and this makes everything feel very pythonic - all data structures will automatically be converted from native R to native Python and numpy types and back.
- V 0.9.2 (2019-12-19)
Replaced deprecated numpy.fromstring with numpy.frombuffer
Refactored exception hierarchy
- V 0.9.1 (2017-05-19)
Removed a bug on some Python3 versions
Added proper support for S4 objects (thanks to flying-sheep)
Added support for Python3 unitests on travis (thanks to flying-sheep)
- V 0.9.0 (2016-04-11)
Full support for data objects larger than 2**24 bytes
Maximum size of message sent to Rserv can now be 2**64 bytes
- V 0.8.4 (2015-09-06)
fixed missing requirements.txt in MANIFEST.in
fixed bug in installer (setup.py)
- V 0.8.3 (2015-09-04)
Fixed exception catching for Python 3.4 (thanks to eeue56)
Some pep8 cleanups
explicit initialization of a number of instance variables in some classes
cleanup of import statements in test modules
Allow for message sizes greater than 4GB coming from R server
- V 0.8.2 (2015-07-11)
Added support for S4 objects (generated when e.g. creating a db object in R)
- V 0.8.1 (2014-07-17)
Fixed errors in the documentation, updated outdated parts
For unittesting run Rserve on different port from the default 6311 to avoid clashes with regular Rserve running on the same server
Fixed but when passing a R-function as argument to a function call (e.g. to sapply), added unittest for this
- V 0.8.0 (2014-06-26)
Added support for remote shutdown of Rserve (thanks to Uwe Schmitt)
Added support for Out-Of-Bounds (OOB) messages (thanks to Philipp alias flying-sheep)
- V 0.7.3 (2013-08-01)
Added missing MANIFEST.in to produce a complete tgz package (now includes docs etc)
Fixed bug on x64 machines when handling integers larger than 2**31
- V 0.7.2 (2013-07-19)
Tested with Python 3.3.x, R 3.0.1 and Rserve 1.7.0
Updated documentation accordingly
Code cleanup for pep8 (mostly)
Marked code as production stable
- V 0.7.1 (2013-06-23)
Added link to new GitHub repository
fixed URL to documentation
- V 0.7.0 (2013-02-25)
Fixed problem when receiving very large result sets from R (added support for XT_LARGE header flag)
Correctly translate multi-dimensional R arrays into numpy arrays (preserve axes the right way) Removed ‘arrayOrder’ keyword argument as a consequence. THIS IS AN API CHANGE - PLEASE CHECK AND ADAPT YOUR CODE, ESPECIALLY IF YOU USE MULTI-DIM ARRAYS!!
Support for conn.voidEval and conn.eval and new ‘defaultVoid’-kw argument in the connect() function
Fixed bug in receiving multi-dimensional boolean (logical) arrays from R
Added support for multi-dimensional string arrays
added support for XT_VECTOR_EXPR type generated e.g. via “expression()” in R (will return a list with the expression content as list content)
windows users can now connect to localhost by pyRserve.connect() (omitting ‘localhost’ parameter)
- V 0.6.0 (2012-06-25)
support for Python3.x
Python versions <= 2.5 no more supported (due to Py3 support)
support for unicode strings in Python 2.x
full support complex numbers, partial support for 64bit integers and arrays
suport for Fortran-style ordering of numpy arrays
elements of single-item arrays are now translated to native python data types
much improved documentation
better unit test coverage
usage of the deprecated conn(<eval-string>) is no more possible
pyRserve.rconnect() now also removed
- V 0.5.2 (2011-12-02)
Fixed problem with 32bit integers being mistakenly rendered into 64bit integers on 64bit machines
- V 0.5.1 (2011-11-22)
Fixed improper DeprecationWarning when evaluating R statements via conn.r(…)
- V 0.5 (2011-10-03)
Renamed pyRserve.rconnect() to pyRserve.connect(). The former still works but shows a DeprecationWarning
String evaluation should now only be executed on the namespace directly, not on the connection object anymore. The latter still works but shows a DeprecationWarning.
New kw argument atomicArray=True added to pyRserve.connect() for preventing single valued arrays from being converted into atomic python data types.
- V 0.4 (2011-09-20)
Added support for nested function calls. E.g. conn.r.t.test( ….) now works.
Proper support for boolean variables and vectors
- V 0.3 (2010-06-08)
Added conversion of more complex R structures into Python
Updated documentation (installation, manual)
V 0.2 (2010-03-19) Fixed rendering of TaggedArrays
V 0.1 (2010-01-10) Initial version
- This package has been mainly developed under Linux, and hence should run on all standard unix platforms, as well
as on Mac OS X. pyRserve has also been successfully used on Win32 machines. Unittests have been used on the Linux and Mac OS X side, however they might just work fine for Win32.
It has been tested run with Python 2.7.x, 3.6, 3.7, and 3.8
The latest development has been tested with R versions 3.0.1 and 3.6.2, and Rserve 1.8.0 and 1.8.6.
pyRserve has been written by Ralph Heinkel (www.ralph-heinkel.com) and is released under MIT license.
From your unix/windows command line run:
For manual installation download the tar.gz or zip package. After unpacking, cd into the pyRserve directory and run python setup.py install from the command line.
Source Code repository
pyRserve is now hosted on GitHub at https://github.com/ralhei/pyRserve.
Documentation can be found at http://packages.python.org/pyRserve/.
For discussion of pyRserve issues and getting help please use the Google newsgroup available at http://groups.google.com/group/pyrserve.
Authentication is implemented in Rserve but not yet in pyRserve
Release history Release notifications | RSS feed
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Hashes for pyRserve-0.9.2-py2.py3-none-any.whl