R help - Nov 2006 - Speeding up small functions

Greetings list,

In my code, I have a few small functions that are called very very  
frequently. An example of one such function is the following :

sigmoid<-function(x) 1/(1+exp(-x))

Now, is there anyway to make this go faster? For example, in C++ we  
could make it inline. Is there a corresponding feature in R?

Cheers,
Wee-Jin

Try a test in which you replace the function calls with inline
code and compare the two speed-wise.  You may very well
find that the difference is inconsequential.

On 11/19/06, Wee-Jin Goh <wjgoh at brookes.ac.uk>
wrote:
> Greetings list,
>
> In my code, I have a few small functions that are called very very
> frequently. An example of one such function is the following :
>
> sigmoid<-function(x) 1/(1+exp(-x))
>
> Now, is there anyway to make this go faster? For example, in C++ we
> could make it inline. Is there a corresponding feature in R?
>
> Cheers,
> Wee-Jin
>
> ______________________________________________
> R-help at stat.math.ethz.ch mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
>

Wee-Jin,

The other option that you have is to set up your function as an expression
and then evaluate the expression for each new value of x.  This might be
faster in some cases.

sigmoid.fun <- function(x) 1/(1+exp(-x))
sigmoid.expr <- expression( 1/(1+exp(-x)) )

x <- runif(10^6)

# non-vectorized function
system.time(
    for(i in seq(along=x)) sigmoid.fun(x[i])
)
[1] 6.76 0.01 7.13   NA   NA

# vectorized function
system.time(
    sigmoid.fun(x)
)
[1] 0.56 0.00 0.59   NA   NA

# vectorized expression
system.time(
    eval(sigmoid.expr)
) 
[1] 0.37 0.01 0.39   NA   NA

-Christos

Christos Hatzis, Ph.D.
Nuvera Biosciences, Inc.
400 West Cummings Park
Suite 5350
Woburn, MA 01801
Tel: 781-938-3830
www.nuverabio.com
 


-----Original Message-----
From: r-help-bounces at stat.math.ethz.ch
[mailto:r-help-bounces at stat.math.ethz.ch] On Behalf Of Wee-Jin Goh
Sent: Sunday, November 19, 2006 4:26 PM
To: r-help at stat.math.ethz.ch
Subject: [R] Speeding up small functions

Greetings list,

In my code, I have a few small functions that are called very very
frequently. An example of one such function is the following :

sigmoid<-function(x) 1/(1+exp(-x))

Now, is there anyway to make this go faster? For example, in C++ we could
make it inline. Is there a corresponding feature in R?

Cheers,
Wee-Jin

______________________________________________
R-help at stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.