Writing R packages in an easier way?

"Dr. Michael Wolf"<m-wolf at muenster.de>  11/02/2011 07:52>>>

is there an easier way to write R packages for the own use - without

RTools and TeX?

Installing Rtools is not hard, and doesn't have to happen often; the
hardest bit in Windows is making sure that the requisite executables are
on the path, and that just involves adding the directory names to the
path environment variable. If I understand you, the problem is the time
spent hacking about in the .Rd help files. That can certainly be
simplified - eliminated, in fact.

Use package.skeleton() once you have a good starting set of functions
and data in R. That creates all the necessary directories, creates
skeleton (but valid) .Rd files, and exports your functions and data
objects for you. You can then edit the code directly, use RCMD check to
check the package (useful anyway) and use RCMD build to build it. (In
fact if all you want is the zip, you can - or at least could - zip the
package directory created by RCMD check).

S Ellison


*******************************************************************
This email and any attachments are confidential. Any use...{{dropped:8}}

On 11.02.2011 13:38, S Ellison wrote:

"Dr. Michael Wolf"<m-wolf at muenster.de> 11/02/2011 07:52>>>

is there an easier way to write R packages for the own use - without

RTools and TeX?

Installing Rtools is not hard, and doesn't have to happen often; the
hardest bit in Windows is making sure that the requisite executables are
on the path, and that just involves adding the directory names to the
path environment variable. If I understand you, the problem is the time
spent hacking about in the .Rd help files. That can certainly be
simplified - eliminated, in fact.

Use package.skeleton() once you have a good starting set of functions
and data in R. That creates all the necessary directories, creates
skeleton (but valid) .Rd files, and exports your functions and data
objects for you. You can then edit the code directly, use RCMD check to
check the package (useful anyway) and use RCMD build to build it. (In
fact if all you want is the zip, you can - or at least could - zip the
package directory created by RCMD check).

Actually, just say

R CMD INSTALL --build package

which will generate the zip in a supported way.

Uwe Ligges

S Ellison


*******************************************************************
This email and any attachments are confidential. Any use...{{dropped:8}}

"Dr. Michael Wolf" <m-wolf at muenster.de> 11/02/2011 07:52 >>>

is there an easier way to write R packages for the own use - without

RTools and TeX?

Installing Rtools is not hard, and doesn't have to happen often; the
hardest bit in Windows is making sure that the requisite executables are
on the path, and that just involves adding the directory names to the
path environment variable. If I understand you, the problem is the time

Dear collegues,

thanks for your helpfull and persuasive comments. I see that all of you
propose to work with the official method building R packages. The code of
importing functions which Barry Rowlingson posted to the forum is very
interesting and perhaps I can use this for solving other problems.
I'm thinking about a monitoring project with R in the center of my working.
Therefore I need help files for describing my programming code.

In the consequence of this I have to accept that using the official R way of
writing a package will be the best in the long run - even it will take some
time to me especially to write the help files. SO I will reactivate my
RTools and TeX!

Best regards

Dr. Michael Wolf
(m-wolf at muenster.de)


Am 11.02.2011 18:49, schrieb Uwe Ligges:

On 11.02.2011 13:38, S Ellison wrote:

"Dr. Michael Wolf"<m-wolf at muenster.de> 11/02/2011 07:52>>>

is there an easier way to write R packages for the own use - without

RTools and TeX?

Installing Rtools is not hard, and doesn't have to happen often; the
hardest bit in Windows is making sure that the requisite executables are
on the path, and that just involves adding the directory names to the
path environment variable. If I understand you, the problem is the time
spent hacking about in the .Rd help files. That can certainly be
simplified - eliminated, in fact.

Use package.skeleton() once you have a good starting set of functions
and data in R. That creates all the necessary directories, creates
skeleton (but valid) .Rd files, and exports your functions and data
objects for you. You can then edit the code directly, use RCMD check to
check the package (useful anyway) and use RCMD build to build it. (In
fact if all you want is the zip, you can - or at least could - zip the
package directory created by RCMD check).

Actually, just say

R CMD INSTALL --build package

which will generate the zip in a supported way.

Uwe Ligges

S Ellison


*******************************************************************
This email and any attachments are confidential. Any use...{{dropped:8}}