use only cmq for hpc

hpc/README.md

@@ -1,41 +1,4 @@
 High performance computing / LSF jobs
 =====================================
 
-Perform function calls in LSF jobs. The module will start `n_jobs` LSF jobs (workers) with the
-memory requirements specified, and then send `<i>` function calls to those workers.
-
-This is done entirely on the network and without temporary files (unless `log_worker=TRUE`),
-so there is no strain on the file system apart from starting up R once per LSF job.
-
-The module also performs load-balancing, i.e. workers that get their jobs done faster will also
-receive more function calls to work on. This is especially useful if not all calls
-return after the same time, or one worker has a high load. For long running jobs use `n_jobs=<i>`.
-
-It is based upon the [rzmq package](https://github.com/armstrtw/rzmq) and the
-[ZeroMQ library](http://zeromq.org/) that is also used for workers in IPython.
-
-It also uses the [ulimit package](https://github.com/krlmlr/ulimit).
-
-The function supplied **must be self-sufficient**, i.e. load libraries and scripts.
-
-### Custom setup
-
-Currently, only LSF is supported as a scheduler. Adding others should be simple, but will
-only be implemented if there is a need for it.
-
-If not at EBI, you may need to adjust the `LSF.tmpl` file according to your needs,
-especially for the queue and custom resources.
-
-### `Q()`
-
-```r
-hpc = import('hpc')
-s = function(x) x
-hpc$Q(s, x=c(1:3), n_jobs=1) # list(1,2,3)
-```
-
-```r
-t = function(x) sum(x)
-a = matrix(3:6, nrow=2)
-hpc$Q(t, a, n_jobs=1) # splits a by columns: list(7, 11)
-```
+This module now uses the [`clustermq`](https://github.com/mschubert/clustermq) package.

hpc/__init__.r

@@ -1 +1 @@
-export_submodule('./master')
+Q = clustermq::Q