# Hi All,
#
# I need to solve a somewhat complex equation at many parameter
values for
# a number of different parameters.
# A simplified version of the equation is: 0= (d1/(h1^2))-(h2*(d2^2))
# I'd like to solve it across a parameter space of d1 and d2, holding
# h1 and h2 constant.
# It seems that uniroot() can do it, but I don't see how to
vectorize it.
# This works just fine:
R12 = function(d1, d2, h1, h2) (d1/(h1^2))-(h2*(d2^2))
uniroot(R12, interval=c(-100,100), tol=0.0001, d2=1, h1=1, h2=1)
# but to explore the entire parameter space it seems I would need a
series of loops
R12 = function(d1, d2, h1, h2) (d1/(h1^2))-(h2*(d2^2))
df1=data.frame(h1=numeric(0), h2=numeric(0),d1=numeric(0),d2=numeric(0))
h1=1
h2=1
for(i in 1:5) {
output=uniroot(R12, interval=c(-100,100), tol=0.0001, d2=i, h1=1, h2=1)
d2=i
df1[i,"h1"]=h1
df1[i,"h2"]=h2
df1[i,"d2"]=i
df1[i,3]=output$root
}
df1
# what I'd really like to do is something like:
R12 = function(d1, d2, h1, h2) (d1/(h1^2))-(h2*(d2^2))
uniroot(R12, interval=c(-100,100), tol=0.0001, d2=c(1:5), h1=c(1:5),
h2=c(1:5))
# but R rejects the multiple values for d2.
# I could use a numerical solver such as XPPAUT, but would rather be
able to do it all in R.
# Any thoughts on vectorizing this would be greatly appreciated.
# Thanks! -Dan
--
Daniel E. Bunker
Associate Coordinator - BioMERGE
Post-Doctoral Research Scientist
Columbia University
Department of Ecology, Evolution and Environmental Biology
1020 Schermerhorn Extension
1200 Amsterdam Avenue
New York, NY 10027-5557
212-854-9881
212-854-8188 fax
deb37ATcolumbiaDOTedu
