| Version: | 0.3.21 |
| Date: | 2025-01-08 |
| Title: | Functions to Inline C, C++, Fortran Function Calls from R |
| Imports: | methods |
| Suggests: | Rcpp, tinytest |
| Description: | Functionality to dynamically define R functions and S4 methods with 'inlined' C, C++ or Fortran code supporting the .C and .Call calling conventions. |
| License: | LGPL-2 |LGPL-2.1 |LGPL-3 [expanded from: LGPL] |
| Copyright: | Oleg Sklyar, 2005-2010 and other authors per their commits |
| LazyLoad: | yes |
| URL: | https://github.com/eddelbuettel/inline,https://dirk.eddelbuettel.com/code/inline.html |
| BugReports: | https://github.com/eddelbuettel/inline/issues |
| NeedsCompilation: | no |
| Packaged: | 2025-01-08 22:04:33 UTC; edd |
| Author: | Oleg Sklyar [aut], Dirk Eddelbuettel |
| Maintainer: | Dirk Eddelbuettel <edd@debian.org> |
| Repository: | CRAN |
| Date/Publication: | 2025-01-09 13:00:02 UTC |
Functions to Inline C, C++, Fortran Function Calls from R
Description
Functionality to dynamically define R functions and S4 methods with 'inlined' C, C++ or Fortran code supporting the .C and .Call calling conventions.
Maintainer
Dirk Eddelbuettel <edd@debian.org>
Author(s)
Oleg Sklyar [aut], Dirk Eddelbuettel [aut, cre] (<https://orcid.org/0000-0001-6419-907X>), Romain Francois [aut] (<https://orcid.org/0000-0002-2444-4226>), Mike Smith [ctb], Duncan Murdoch [ctb], Karline Soetaert [ctb] (<https://orcid.org/0000-0003-4603-7100>), Johannes Ranke [ctb] (<https://orcid.org/0000-0003-4371-6538>)
See Also
Inline C, C++, Fortran function calls from R
Description
Functionality to dynamically define R functions and S4 methods with in-lined C,C++ or Fortran code supporting .C and .Call calling conventions.
Usage
cfunction(sig=character(), body=character(), includes=character(), otherdefs=character(), language=c("C++", "C", "Fortran", "F95", "ObjectiveC", "ObjectiveC++"), verbose=FALSE, convention=c(".Call", ".C", ".Fortran"), Rcpp=FALSE, cppargs=character(), cxxargs=character(), libargs=character(), dim=NULL, implicit=NULL, module=NULL, name=NULL) ## S4 methods for signatures # f='character', sig='list', body='list' # f='character', sig='character', body='character' setCMethod(f, sig, body, ...) ## Further arguments: # setCMethod(f, sig, body, includes="", otherdefs="", cpp=TRUE, # verbose=FALSE, where=topenv(.GlobalEnv), ...)Arguments
f | A single character value if |
sig | A match of formal argument names for the function with thecharacter-string names of corresponding classes. Alternatively,a named list of such character vectors. The names of the list elements willbe used as function names (see example). If |
body | A character vector with C, C++ or Fortran code omitting functiondeclaration (only the body, i.e. in case of C starting after the functionopening curly bracket and ending before the closing curly bracket,brackets excluded). In case of |
includes | A character vector of additional includes and preprocessorstatements etc that will be put between the R includes and the user function(s). |
otherdefs | A characted vector with the code for any further definitions offunctions, classes, types, forward declarations, namespace usage clauses etcwhich is inserted between the includes and the declarations of the functionsdefined in |
language | A character value that specifies the source language of theinline code. The possible values for |
verbose | If |
convention | Which calling convention to use? See the Details section. |
Rcpp | If |
cppargs | Optional character vector of tokens to be passed tothe compiler via the |
cxxargs | Optional character vector of tokens to be passed tothe compiler via the |
libargs | Optional character vector of tokens to be passed to thecompiler via the |
dim | Optional character vector defining the dimensionality of thefunction arguments. Of same length as |
implicit | A character vector defining the implicit declaration inFortran or F95; the default is to use the implicit typing rules for Fortran,which is |
module | Name(s) of any modules to be used in the |
name | Function name to be used in the code. Only used if |
... | Reserved. |
Details
To declare multiple functions in the same library one can usesetCMethodsupplying lists of signatures and implementations. In this case, provide asmany method names inf as you define methods. Avoid clashes when selectingnames of the methods to declare, i.e. if you provide the same name several timesyou must ensure that signatures are different but can share the same generic!
The source code in thebody should not include the header or"front-matter" of the function or the close, e.g. in C or C++ itmust start after the C-function opening curly bracket and end beforethe C-function closing curly bracket, brackets should not beincluded. The header will be automatically generated from the R-signatureargument. Arguments will will carry the same name as used in the signature,so avoid variable names that are not legal in the target language(e.g. names with dots).
C/C++: Ifconvention == ".Call" (the default), the.Call mechanismis used and its result is returned directly as the result of the call of thegenerated function. As the last line of the generated C/C++ code areturn R_NilValue; is added in this case and a warning is generatedin case the user has forgotten to provide a return value. To suppress thewarning and still return NULL, addreturn R_NilValue; explicitly.
Special care is needed with types, memory allocation and protection– exactly the same as if the code was not inline: see theWriting R Extension manual for information on.Call.
Ifconvention == ".C" orconvention == ".Fortran", the.C or.Fortran mechanism respectively isused, and the return value is a list containing all arguments.
Attached R includes includeR.h for".C", andadditionallyRdefines.h andR_ext\Error.h for".Call".
Value
Ifsig is a single character vector,cfunction returns a singlefunction; if it is a list, it returns a list of functions.
setCMethod declares new methods with given names and signatures andreturns invisibleNULL.
Author(s)
Oleg Sklyar, Duncan Murdoch, Mike Smith, Dirk Eddelbuettel
See Also
Foreign Function Interface
Examples
x <- as.numeric(1:10)n <- as.integer(10)## Not run: ## A simple Fortran example - n and x: assumed-size vectorcode <- " integer i do 1 i=1, n(1) 1 x(i) = x(i)**3"cubefn <- cfunction(signature(n="integer", x="numeric"), code, convention=".Fortran")print(cubefn)cubefn(n, x)$x## Same Fortran example - now n is one numbercode2 <- " integer i do 1 i=1, n 1 x(i) = x(i)**3"cubefn2 <- cfunction(signature(n="integer", x="numeric"), implicit = "none", dim = c("", "(*)"), code2, convention=".Fortran")cubefn2(n, x)$x## Same in F95, now x is fixed-size vector (length = n)code3 <- "x = x*x*x"cubefn3 <- cfunction(sig = signature(n="integer", x="numeric"), implicit = "none", dim = c("", "(n)"), code3, language="F95")cubefn3(20, 1:20)print(cubefn3)## Same example in Ccode4 <- " int i; for (i = 0; i < *n; i++) x[i] = x[i]*x[i]*x[i];"cubefn4 <- cfunction(signature(n="integer", x="numeric"), code4, language = "C", convention = ".C")cubefn4(20, 1:20)## Give the function in the source code a namecubefn5 <- cfunction(signature(n="integer", x="numeric"), code4, language = "C", convention = ".C", name = "cubefn")code(cubefn5)## End(Not run) ## use of a module in F95modct <- "module modctsdouble precision, parameter :: pi = 3.14159265358979double precision, parameter :: e = 2.71828182845905end"getconstants <- "x(1) = pix(2) = e"cgetcts <- cfunction(getconstants, module = "modcts", implicit = "none", includes = modct, sig = c(x = "double"), dim = c("(2)"), language = "F95")cgetcts(x = 1:2)print(cgetcts)## Use of .C convention with C code## Defining two functions, one of which calls the othersigSq <- signature(n="integer", x="numeric")codeSq <- " for (int i=0; i < *n; i++) { x[i] = x[i]*x[i]; }"sigQd <- signature(n="integer", x="numeric")codeQd <- " squarefn(n, x); squarefn(n, x);"fns <- cfunction( list(squarefn=sigSq, quadfn=sigQd), list(codeSq, codeQd), convention=".C")squarefn <- fns[["squarefn"]]quadfn <- fns[["quadfn"]]squarefn(n, x)$xquadfn(n, x)$x## Alternative declaration using 'setCMethod'setCMethod(c("squarefn", "quadfn"), list(sigSq, sigQd), list(codeSq, codeQd), convention=".C")squarefn(n, x)$xquadfn(n, x)$x## Use of .Call convention with C code## Multyplying each image in a stack with a 2D Gaussian at a given positioncode <- " SEXP res; int nprotect = 0, nx, ny, nz, x, y; PROTECT(res = Rf_duplicate(a)); nprotect++; nx = INTEGER(GET_DIM(a))[0]; ny = INTEGER(GET_DIM(a))[1]; nz = INTEGER(GET_DIM(a))[2]; double sigma2 = REAL(s)[0] * REAL(s)[0], d2 ; double cx = REAL(centre)[0], cy = REAL(centre)[1], *data, *rdata; for (int im = 0; im < nz; im++) { data = &(REAL(a)[im*nx*ny]); rdata = &(REAL(res)[im*nx*ny]); for (x = 0; x < nx; x++) for (y = 0; y < ny; y++) { d2 = (x-cx)*(x-cx) + (y-cy)*(y-cy); rdata[x + y*nx] = data[x + y*nx] * exp(-d2/sigma2); } } UNPROTECT(nprotect); return res;"funx <- cfunction(signature(a="array", s="numeric", centre="numeric"), code)x <- array(runif(50*50), c(50,50,1))res <- funx(a=x, s=10, centre=c(25,15))if (interactive()) image(res[,,1])## Same but done by registering an S4 methodsetCMethod("funy", signature(a="array", s="numeric", centre="numeric"), code, verbose=TRUE)res <- funy(x, 10, c(35,35))if (interactive()) { x11(); image(res[,,1]) }inline C++ function
Description
Functionality to dynamically define an R function with inlined C++ codeusing the.Call calling convention.
Thercpp() wrapper sets the plugin to the “Rcpp” valuesuitable for usingRcpp.
Usage
cxxfunction(sig = character(), body = character(), plugin = "default", includes = "", settings = getPlugin(plugin), ..., verbose = FALSE)rcpp(..., plugin="Rcpp")Arguments
sig | Signature of the function. A named character vector |
body | A character vector with C++ code to include in the body of the compiled C++ function |
plugin | Name of the plugin to use. See |
includes | User includes, inserted after the includes provided by the plugin. |
settings | Result of the call to the plugin |
... | Further arguments to the plugin |
verbose | verbose output |
Value
A function
See Also
Examples
## Not run: # default pluginfx <- cxxfunction(signature(x = "integer", y = "numeric"), "return Rf_ScalarReal(INTEGER(x)[0] * REAL(y)[0]);")fx(2L, 5)# Rcpp pluginif (requireNamespace("Rcpp", quietly=TRUE)) { fx <- cxxfunction(signature(x = "integer", y = "numeric"), "return wrap( as<int>(x) * as<double>(y));", plugin = "Rcpp" ) fx(2L, 5) ## equivalent shorter form using rcpp() fx <- rcpp(signature(x = "integer", y = "numeric"), "return wrap(as<int>(x) * as<double>(y));")}# RcppArmadillo pluginif (requireNamespace(RcppArmadillo)) { fx <- cxxfunction(signature(x = "integer", y = "numeric"), "int dim = as<int>(x); arma::mat z = as<double>(y) * arma::eye<arma::mat>(dim, dim); return wrap(arma::accu(z));", plugin = "RcppArmadillo") fx(2L, 5)}## End(Not run)Retrieve the dynamic library (or DLL) associated with a package of a function generated by cfunction
Description
ThegetDynLib function retrieves the dynamic library (or DLL) associated with a package or with a function generated bycfunction
Methods
signature(x = "CFunc")Retrieves the dynamic library associated with the function generated by
cfunction. The library is dynamically loaded if necessary.signature(x = "CFuncList")Retrieves the dynamic library associated with a set of functions generated by
cfunction. The library is dynamically loaded if necessary.signature(x = "character")Retrieves the dynamic library of the given name. This typically refers to package names, but can be any name of the list returned by
getLoadedDLLs
See Also
Examples
## Not run: getDynLib( "base" )f <- cfunction( signature() , "return R_NilValue ;" )getDynLib( f )## End(Not run)Internals
Description
Aliases required for 'R CMD check' but those no-one will ever search for.Here to prevent the mess of the index. The corresponding items have helpaliases without method signature!
Generate the skeleton of a package
Description
Generate the skeleton of a package
Methods
signature(name = "ANY", list = "ANY")Standard method. See
package.skeletonsignature(name = "character", list = "CFunc")Method for a single generated by
cfunctionorcxxfunctionsignature(name = "character", list = "CFuncList")Method for a set functions generated by
cfunctionorcxxfunction
Examples
## Not run: fx <- cxxfunction(signature(x = "integer", y = "numeric"), "return Rf_ScalarReal( INTEGER(x)[0] * REAL(y)[0]);")package.skeleton("foo", fx)functions <- cxxfunction(list(ff = signature(), gg = signature(x = "integer", y = "numeric")), c("return R_NilValue ;", "return Rf_ScalarReal(INTEGER(x)[0] * REAL(y)[0]);"))package.skeleton("foobar", functions)## End(Not run)Plugin system for cxxfunction
Description
cxxfunction uses a plugin system to assembly the code that it compiles. These functions allow to register and get plugins by their name.
Usage
getPlugin(name, ...)registerPlugin(name, plugin)Arguments
name | name of the plugin. |
... | Further argments to pass to the plugin. |
plugin | plugin function. |
Details
plugins are functions that return a list with :
- includes
mandatory. it is included at the top of the compiled file by
cxxfunction- body
optional. a function that takes one argument (the body of the c++ function) and returneda modified version of the body. The "Rcpp" plugin uses this to surround the code with the
BEGIN_RCPPandEND_RCPPmacros- LinkingTo
optional. character vector containing the list of packagesthat the code needs to link to. This adds the include path of the given packages. The "Rcpp" and "RcppArmadillo" plugins use this.
- env
optional. named list of environment variables. For example, the "Rcpp" plugin uses this to add Rcpp user library to the
PKG_LIBSenvironmentvariable.
plugins can be manually registered using theregisterPluginfunction. Alternatively, a package may supply an inline plugin implicitely by defining a function calledinlineCxxPlugin, which does not necessarily need to be exported from the namespace of the package.
Known packages implementing this scheme includeRcpp andRcppArmadillo.
Value
getPlugin retrieves the plugin and invokes it with the... arguments
registerPlugin does not return anything.
See Also
Examples
## Not run: getPlugin( "Rcpp" )## End(Not run)Printing, reading and writing compiled function objects
Description
moveDLL moves the DLL used by a compiled function to a user definedlocation.
writeCFunc saves aCFunc object after the DLL has been moved tothe desired location usingmoveDLL.
readCFunc reads aCFunc object that has been saved usingwriteCFunc.
Theprint andcode methods respectively print the entireobject or only the code parts.
Usage
moveDLL(x, ...)## S4 method for signature 'CFunc'moveDLL(x, name, directory, unload = FALSE, overwrite = FALSE, verbose = FALSE)writeCFunc(x, file)readCFunc(file)## S4 method for signature 'CFunc'print(x)## S4 method for signature 'CFuncList'print(x)## S4 method for signature 'CFunc'code(x, linenumbers = TRUE)## S4 method for signature 'CFuncList'code(x, linenumbers = TRUE)Arguments
x | A |
name | The base of the file name that the DLL should be moved to. The file nameextension will depend on the operating system used |
directory | The directory that the DLL should be written to |
unload | In case the new path constructed from |
overwrite | In case there is a file at the new path constructed from |
verbose | Should we print a message stating where the DLL was copiedif the operation was successful? |
file | The file path for writing and reading the object generated by |
linenumbers | If |
... | May be used in future methods |
Details
If you move the DLL to a user defined location withmoveDLL, this willkeep an on-disk copy of the DLL which will prevent it from being lost atsession termination - unless written to the sessiontempdir.Saving and reloading theCFunc object with standard tools likesave orsaveRDS will still loose the pointer tothe DLL. However, when the DLL has been moved usingmoveDLL,CFunc objects can be saved bywriteCFunc and restored byreadCFunc.
Value
FunctionreadDynLib returns aCFunc object.
FunctionwriteDynLib returns the name of the.CFunc file thatwas created.
Note
The code of a
CFuncorCFuncListobjectxcan be extracted(rather than printed), using:x@code.To write the code to a file (here called
"fn"), without thenew-line character"\n":write (strsplit(x, "\n")[[1]], file = "fn")
Author(s)
Karline Soetaert and Johannes Ranke
See Also
Examples
x <- as.numeric(1:10)n <- as.integer(10)code <- " integer i do 1 i=1, n(1) 1 x(i) = x(i)**3"cubefn <- cfunction(signature(n="integer", x="numeric"), code, convention=".Fortran")code(cubefn)cubefn(n, x)$x## Not run: # The following code is exempted from the automated tests of example code, as# it writes to the users home directory.# The following writes the DLL, e.g. cubefn.so on Linux/Unix or cubefn.dll# on WindowsmoveDLL(cubefn, name = "cubefn", directory = "~")path <- file.path("~", "cubefn.rda")writeCFunc(cubefn, path)rm(cubefn)# Now you can start a fresh R session and load the functionlibrary(inline)path <- file.path("~", "cubefn.rda")cfn <- readCFunc(path)cfn(3, 1:3)$x## End(Not run)