Next: Small example, Previous: Installation, Up: Starting up
Once installed, the module can be imported with:
>>> from rpy import *
If an error occurs, refer to the section TROUBLESHOOTING on the README file.
The module rpy
imports a Python object named r
, from
which all the R functions and objects can be retrieved, see R objects look up. This module also implements a new Python type:
Robj
, which represents an arbitrary R object, see Robj type. For example:
>>> r.wilcox_test <Robj object at 0x8a9e120>
is the R function wilcox.test
which computes the Wilcoxon
statistical test. An object of type Robj
is always callable as
long as the corresponding R object is:
>>> r.wilcox_test([1,2,3], [4,5,6]) {'p.value': 0.10000000000000001, 'statistic': {'W': 0.0}, 'null.value': {'mu': 0.0}, 'data.name': 'c(1, 2, 3) and c(4, 5, 6)', 'alternative': 'two.sided', 'parameter': None, 'method': 'Wilcoxon rank sum test'}
The arguments are translated automatically to R objects and the return value is translated back to Python, when this is possible (see Conversion system). This autoconversion can be customized at several levels or disabled at all.
Objects of type Robj
also support keyword arguments, in the
same way as normal Python functions:
>>> r.seq(1, 3, by=0.5) [1.0, 1.5, 2.0, 2.5, 3]
The module rpy
defines a new exception type derived from the
base class Exception
, called RException
, see
R exceptions. When any kind of error in the R interpreter
occurs, an exception of this type is raised:
>>> r.plot() Traceback (most recent call last): File "<stdin>", line 1, in ? rpy.RException: Error in function (x, ...) : Argument "x" is missing, with no default