[R] flexible interfaces

Paul Gilbert pgilbert at bank-banque-canada.ca
Fri Dec 4 22:26:06 CET 1998

>In contrast, free packages should strive to support flexible interfaces
>between applications, although getting that to work smoothly outside unix
>may not be feasible.

We have a interface which provides a generic way for applications to read and
write objects to a database. (This is not a query interface.) I use it for
transferring time series data between S/R and Fame, but there is nothing which
makes it especially specific to S or R or Fame. The protocol is not specific to
time series, but the data format we use is. It handles endian problems and uses
RPC so it works over the Internet. It communicates directly and does not write
intermediate files. I work in Unix but a commercial vendor has  implemented a
Windows version and also a version for Matlab. (I have not used these.) There
should be no cross-platform problem. Details and the code are available at
<http://www.bank-banque-canada.ca/pgilbert>. (I have a version of the code which
works with R but it is not on the web site yet. If you want that please contact

There appears to be some overlap between what we have done and HDF or NetCDF but
I am not too familiar with these. (If someone knowledgeable would be prepared to
brief me I would appreciate that.)

I expect someday CORBA will be the way to go. There is an ongoing effort to by
an OMG "Special Interest Group/ Domain Task Force" to define statistical objects
and they currently have an RFP/RFI for which I believe they are hoping to have
reponses by Dec. 19. (A few more details are attached below. I can provide more
information if anyone is interested, please email me directly rather than
through the list.) I believe this is being co-ordinated with the UN/EDIFACT
"generalized statistical message" (GESMES) effort.

Paul Gilbert

 Descriptions and schedules for all RFPs and RFIs in progress are
available for viewing on the OMG Web site at

r-help mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-help-request at stat.math.ethz.ch

More information about the R-help mailing list