add support for multi-threading

[ajaech]

A new command line argument controls how many threads to use. A thread
pool is created and the main work of updating the data structures after
a cluster merge is split equally among all the threads in the pool. The
clustering can be more than 10 times as fast using multiple threads,
depending on how many cores on the CPU.

[ajaech]

I made a few changes to make the clustering run faster by using multiple threads. I've been testing it for a couple of weeks. It gives exactly the same clustering as before except it is much faster.

[percyliang]

Thank you - this will be very helpful to people!

-Percy

On Wed, Mar 26, 2014 at 10:30 AM, ajaech notifications@github.com wrote:

I made a few changes to make the clustering run faster by using multiple
threads. I've been testing it for a couple of weeks. It gives exactly the
same clustering as before except it is much faster.

Reply to this email directly or view it on GitHubhttps://github.com//pull/9#issuecomment-38713957
.

[arianhosseini]

what is the command line argument -t or ... ?! and how many threads are safe to use?!

[ajaech]

Are you asking what the command line option is to set the number of
threads? I think it is --threads. See here
https://github.com/percyliang/brown-cluster/blob/master/wcluster.cc#L63.

It is safe to use as many threads as you want but you will probably not get
any benefit to using more threads than there are cores on your cpu.

On Tue, Nov 24, 2015 at 11:56 AM, Arian notifications@github.com wrote:

what is the command line argument -t or ... ?! and how many threads are
safe to use?!

—
Reply to this email directly or view it on GitHub
#9 (comment)
.

[arianhosseini]

thanks, I think adding the command line option in the Read Me would help others to understand how to use this feature.

[ajaech]

When you run the script with no arguments it should print out a list of all
the arguments along with short descrptions.

On Tue, Nov 24, 2015 at 12:05 PM, Arian notifications@github.com wrote:

thanks, I think adding the command line option in the Read Me would help
others to understand how to use this feature.

—
Reply to this email directly or view it on GitHub
#9 (comment)
.