We've been running R/qtl for several months now without any issues
except time.
For our data the ripple step executes for weeks, so clearly we want to
do it in parallel.
We installed SNOW and mpi on our linux cluster (fedora) and added the
n.cluster parameter to our ripple call:
rip1lik <- ripple( mydata,
chr=1,
window=2,
method="likelihood",
error.prob=0.05,
map.function="kosambi",
maxit=100000,
n.cluster = 8 )
The response is:
rip1lik<-
ripple(mydata,chr=1,window=2,method="likelihood",error.prob=0.05,map.function="kosambi",maxit=100000,n.cluster=8)
974 total orders
[
mpi...@inquiry.unh.edu] match_arg (./utils/args/args.c:122):
unrecognized argument pmi_args
[
mpi...@inquiry.unh.edu] HYDU_parse_array (./utils/args/args.c:140):
argument matching returned error
[
mpi...@inquiry.unh.edu] parse_args (./ui/mpich/utils.c:1387): error
parsing input array
[
mpi...@inquiry.unh.edu] HYD_uii_mpx_get_parameters (./ui/mpich/
utils.c:1438): unable to parse user arguments
Usage: ./mpiexec [global opts] [exec1 local opts] : [exec2 local
opts] : ...
Global options (passed to all executables):
<and a lot more "usage" information>
----------------------------------------------------------------------
Our first effort had yielded an error message that the software could
not find "libmpl.so.1". Our sysadmin then updated the mpich2
installation to one that includes libmpl.so.1 and we got a little
farther. Has anyone seen this problem before? Are there multiple
versions of mpich2 that are different?
Thanks