R has a documentation system that ensures that documentation for code distributed as packages is installed when packages are installed. This documentation can be called and searched from R itself.
Unlike Python docstrings, where the documentation string
can be found in the special attribute
the R documentation lives outside objects in documentation pages.
Each documentation page is associated at minimum one alias, aliases often
corresponding to the name of an R object defined in a package
(function, dataset, etc…).
For example, querying documentation for the R function sum becomes a matter of finding which documentation page has the alias sum, and retrieve that page.
Querying on aliases¶
When working with R, a frequent use case for using the documention is to query on an alias (a function name, a dataset, or a class name) and retrieve the associated documentation.
While the R packaging system will make checks that any given alias is associated with only one page within the same package, it is well possible to have several packages defining a documentation page for the same alias.
With rpy2’s interface to the help system, an easy way to retrive
pages associated with an alias is to
use the function
pages(), which returns a
The documentation for a package is represented with the class
>>> import rpy2.robjects.help as rh >>> base_help = rh.Package('base') >>> base_help.fetch('sum')
A documentation page is represented as an instance of
>>> hp = base_help.fetch('sum')
>>> hp.sections.keys() ('title', 'name', 'alias', 'keyword', 'description', 'usage', 'arguments', 'deta ils', 'value', 'section', 'references', 'seealso')
details ------- This is a generic function: methods can be defined for it directly or via the Summary group generic. For this to work properly, the arguments should be unnamed, and dispatch is on the first argument. If na.rm is FALSE an NA value in any of the arguments will cause a value of NA to be returned, otherwise NA values are ignored. Logical true values are regarded as one, false values as zero. For historical reasons, NULL is accepted and treated as if it were integer(0) .