Great toolbox. Thanks!
You may want to substitute current lines for:
if isnan(s) % user does not enter any seed, use a random seed
myseed=sum(100*clock);
else % user seed to repeat the same computation
myseed=seed;
end
if verLessThan(‘matlab’,’8.0′)
RandStream.setDefaultStream(RandStream(‘mt19937ar’,’seed’,myseed));
else
RandStream.setGlobalStream(RandStream(‘mt19937ar’,’seed’,myseed));
end
in TMVN.m and multi_norm_sample_from.m. Otherwise, modern ML stops with errors.
Great toolbox. Thanks!
You may want to substitute current lines for:
if isnan(s) % user does not enter any seed, use a random seed
myseed=sum(100*clock);
else % user seed to repeat the same computation
myseed=seed;
end
if verLessThan(‘matlab’,’8.0′)
RandStream.setDefaultStream(RandStream(‘mt19937ar’,’seed’,myseed));
else
RandStream.setGlobalStream(RandStream(‘mt19937ar’,’seed’,myseed));
end
in TMVN.m and multi_norm_sample_from.m. Otherwise, modern ML stops with errors.
Thanks for valuable suggestions. Will do it ~