R-alpha: pow_ii

Paul Gilbert · 1997-08-28T16:01:05Z

>> I'm a bit confused about this discussion, but since I don't have f2c on ... >A lot of Fortran code doesn't require any libf2c routines. When routines >are needed the most common ones are the exponentiation routines pow_dd, >pow_ii and pow_di (pow_ri is for single precision reals and so is less >needed in R, which uses double precision). However, *some* fortran code, ... Thomas >From your previous response to my confusion I guessed I would be ok with pow_ii, but that does not seem to be t

2 messages · Paul Gilbert, Thomas Lumley

Original

Paul Gilbert

Thu, Aug 28, 1997 9:01 AM #

...

...

Thomas

with pow_ii, but that does not seem to be the case:

R
...

Random number generator tests ...ok
DSE curvature test 1 ... ld.so.1: /R/bin/R.binary: fatal: relocation error: symbol not found: pow_ii: referenced in /dse/dsefor.so
Killed

[4] /R : nm bin/R.binary | grep pow
[146]	|    244556|     312|FUNC |LOCL |0    |7      |complex_pow
[2557]	|   1462800|       0|FUNC |GLOB |0    |UNDEF  |pow
[2795]	|   1053360|      68|FUNC |GLOB |0    |7      |pow_dd
[2800]	|   1053124|     236|FUNC |GLOB |0    |7      |pow_di
[377]	|   1474036|       4|OBJT |LOCL |0    |15     |powerSymbol
[5] /R : 
[5] /R : R


This leads me to wonder how R calculates powers and exp for complex
numbers. I presume that if I can arrange to call the same routines then
I can avoid the extra burden of having to include parts of libf2c with
my distributed code. Do you know what R uses and can I call these from
fortran?

Paul Gilbert
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
r-devel 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-devel-request@stat.math.ethz.ch
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

Thomas Lumley

Thu, Aug 28, 1997 11:36 AM #

On Thu, 28 Aug 1997, Paul Gilbert wrote:

No, pow_ii doesn't seem to be in there

There is code in src/main/complex.c

This contains routines for powers, division and transcendental functions.
Unfortunately they are all declared static, so you can't use them. 

I don't know if there is any real need to make these static, apart from
tidiness (Anyone?) so it's possible that these could be changed in the
future.

Still, I liked Ross' suggestion of just making a f2c R package


Thomas Lumley
------------------------------------------------------+------
Biostatistics		: "Never attribute to malice what  :
Uni of Washington	:  can be adequately explained by  :
Box 357232		:  incompetence" - Hanlon's Razor  :
Seattle WA 98195-7232	:				   :
------------------------------------------------------------

=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
r-devel 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-devel-request@stat.math.ethz.ch
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-