Internal API

Types

RCall.RObjectType

An RObject is a Julia wrapper for an R object (known as an "S-expression" or "SEXP"). It is stored as a pointer which is protected from the R garbage collector, until the RObject itself is finalized by Julia. The parameter is the type of the S-expression.

When called with a Julia object as an argument, a corresponding R object is constructed.

julia> RObject(1)
RObject{IntSxp}
[1] 1

julia> RObject(1:3)
RObject{IntSxp}
[1] 1 2 3

julia> RObject(1.0:3.0)
RObject{RealSxp}
[1] 1 2 3
source
RCall.SxpType

RCall.jl's type Sxp mirrors the R symbolic expression record SEXPREC in R API. These are represented by a pointer Ptr{S<:Sxp} (which is called SEXP in R API).

source

Methods

Base.isasciiMethod

Determines the encoding of the CharSxp. This is determined by the 'gp' part of the sxpinfo (this is the middle 16 bits).

  • 0x00000200 (bit 1): set of bytes (no known encoding)
  • 0x00000400 (bit 2): Latin-1
  • 0x00000800 (bit 3): UTF-8
  • 0x00004000 (bit 6): ASCII

We only support ASCII and UTF-8.

source
Base.lengthMethod

Sxp methods for length return the R length.

Rf_xlength handles Sxps that are not vector-like and R's "long vectors", which have a negative value for the length member.

source
Base.namesMethod

Returns the names of an R vector, the result is converted to a Julia symbol array.

source
RCall.boundMethod

The R NAMED property, represented by 2 bits in the info field. This can take values 0,1 or 2, corresponding to whether it is bound to 0,1 or 2 or more symbols. See http://cran.r-project.org/doc/manuals/r-patched/R-exts.html#Named-objects-and-copying

source
RCall.getNamespaceMethod

get namespace by name of the namespace. It is safer to be used than findNamespace as it checks bound.

source
RCall.ijulia_setdeviceMethod

Set options for R plotting with IJulia.

The first argument should be a MIME object: currently supported are

  • MIME("image/png") [default]
  • MIME("image/svg+xml")

The remaining arguments (keyword only) are passed to the appropriate R graphics device: see the relevant R help for details.

source
RCall.initEmbeddedRMethod
initEmbeddedR()

This initializes an embedded R session. It should only be called when R is not already running (e.g. if Julia is running inside an R session)

source
RCall.isNAMethod

Check if a value corresponds to R's sentinel NA values. These function should not be exported.

source
RCall.isnaMethod

Check if the ith member of s correspond to R's NA values.

source
RCall.isnaMethod

Check if the members of a vector are NA values. Always return a BitArray.

source
RCall.julia_extptr_callbackMethod

The function called by R .External for Julia callbacks.

It receives a Ptr{ListSxp} containing

  • a pointer to the function itself (Ptr{ExtPtrSxp})
  • a pointer to the Julia function (Ptr{ExtPtrSxp})
  • any arguments (as Ptr{S<:Sxp})
source
RCall.makeNativeSymbolRefMethod

Register a function pointer as an R NativeSymbol. We technically are supposed to use RregisterRoutines. Starting from R 3.4, `RMakeExternalPtrFn` is a part of R API in R 3.4. It is probably safe to such to make the external pointer.

source
RCall.newEnvironmentMethod
newEnvironment([env])

Create a new environment which extends environment env (globalEnv by default).

source
RCall.preserveMethod

Prevent garbage collection of an R object. Object can be released via release.

This is slower than protect, as it requires searching an internal list, but more flexible.

source
RCall.protectMethod

Stack-based protection of garbage collection of R objects. Objects are released via unprotect. Returns the same pointer, allowing inline use.

This is faster than preserve, but more restrictive. Really only useful inside functions.

source
RCall.rcallMethod

Evaluate a function in the global environment. The first argument corresponds to the function to be called. It can be either a FunctionSxp type, a SymSxp or a Symbol.

source
RCall.rcall_pMethod

Evaluate a function in the global environment. The first argument corresponds to the function to be called. It can be either a FunctionSxp type, a SymSxp or a Symbol.

source
RCall.rcopyMethod

rcopy(r) copies the contents of an R object into a corresponding canonical Julia type.

source
RCall.renderMethod

Render an inline R script, substituting invalid "$(Expr(:incomplete, Base.Meta.ParseError("ParseError:\n# Error @ none:1:26\n\" signs for Julia symbols\n# └ ── unterminated string literal", Base.JuliaSyntax.ParseError(Base.JuliaSyntax.SourceFile("\" signs for Julia symbols", 0, "none", 1, [1, 26]), Base.JuliaSyntax.Diagnostic[Base.JuliaSyntax.Diagnostic(26, 25, :error, "unterminated string literal")], :string))))

source
RCall.revalFunction

Evaluate an R symbol or language object (i.e. a function call) in an R try/catch block, returning an RObject.

source
RCall.reval_pMethod

Evaluate an R expression array iteratively. If throw_error is false, the error message and warning will be thrown to stderr.

source
RCall.reval_pMethod

Evaluate an R symbol or language object (i.e. a function call) in an R try/catch block, returning a Sxp pointer.

source
RCall.rimportFunction

Import an R package as a julia module. Sanitizes the imported symbols by default, because otherwise certain symbols cannot be used.

E.g.: PerformanceAnalytics::charts.Bar in R becomes PerformanceAnalytics.charts_Bar in Julia

gg = rimport("ggplot2")
source
RCall.rlangMethod

Create a function call from a function pointer and a list of arguments and return it as an RObject, which can then be evaluated

source
RCall.robjectMethod

robject(x) converts a Julia object x to a corresponding RObject implicitly. Explicit conversions could be called with robject(<R Class>, x).

source
RCall.sexpMethod

sexp(x) converts a Julia object x to a pointer to a corresponding Sxp Object.

source
RCall.sexpMethod

Convert a Ptr{UnknownSxp} to an appropriate Ptr{S<:Sxp}.

source
RCall.sexpMethod

Wrap a Julia object an a R Ptr{ExtPtrSxp}.

We store the pointer and the object in a const Dict to prevent it being removed by the Julia GC.

source
RCall.sexpMethod

Wrap a callable Julia object f an a R ClosSxpPtr.

Constructs the following R code

function(...) .External(juliaCallback, fExPtr, ...)
source
RCall.sexpnumMethod

The SEXPTYPE number of a Sxp

Determined from the trailing 5 bits of the first 32-bit word. Is a 0-based index into the info field of a SxpHead.

source
RCall.unsafe_vecMethod

Represent the contents of a VectorSxp type as a Vector.

This does not copy the contents. If the argument is not named (in R) or otherwise protected from R's garbage collection (e.g. by keeping the containing RObject in scope) the contents of this vector can be modified or could cause a memory error when accessed.

The contents are as stored in R. Missing values (NA's) are represented in R by sentinels. Missing data values in RealSxp and CplxSxp show up as NaN and NaN + NaNim, respectively. Missing data in IntSxp show up as -2147483648, the minimum 32-bit integer value. Internally a LglSxp is represented as Vector{Int32}. The convention is that 0 is false, -2147483648 is NA and all other values represent true.

source
RCall.validate_libRMethod
validate_libR(libR)

Checks that the R library libR can be loaded and is satisfies version requirements.

source

Macros

RCall.@R_strMacro
R"..."

An inline R expression, the result of which is evaluated and returned as an RObject.

It supports substitution of Julia variables and expressions via prefix with $ whenever not valid R syntax (i.e. when not immediately following another completed R expression):

R"glm(Sepal.Length ~ Sepal.Width, data=$iris)"

It is also possible to pass Julia expressions:

R"plot(#50)"

All such Julia expressions are evaluated once, before the R expression is evaluated.

The expression does not support assigning to Julia variables, so the only way retrieve values from R via the return value.

source
RCall.@rimportMacro

Import an R Package as a Julia module. For example,

@rimport ggplot2

is equivalent to ggplot2 = rimport("ggplot2") with error checking.

You can also use classic Python syntax to make an alias: @rimport *package-name* as *shorthand*

@rimport ggplot2 as gg

which is equivalent to gg = rimport("ggplot2").

source
RCall.@rlibraryMacro

Load all exported functions/objects of an R package to the current module. Almost equivalent to

__temp__ = rimport("ggplot2")
using .__temp__
source
RCall.@var_strMacro

Returns a variable named "str". Useful for passing keyword arguments containing dots.

source

Constants

RCall.jtypExtPtrsConstant

Julia types (typically functions) which are wrapped in Ptr{ExtPtrSxp} are stored here to prevent garbage collection by Julia.

source