[R] [R-pkgs] AQ-R 0.2 // realtime messaging.
Ulrich Staudinger
ustaudinger at activequant.com
Wed Dec 5 14:09:42 CET 2012
Hi there,
I am glad to announce AQ-R 0.2 has been successfully built and is
available via install.packages("aqr",
repos="http://R-Forge.R-project.org").
The most important new feature is real-time messaging from within R.
AQ-R 0.2 enables you to send and receive byte[] messages within R
through a STOMP compliant messaging server, such as the ActiveQuant
Master Server. There are various STOMP protocol compliant servers [1].
You can build arbitrarily complex messaging infrastructures, where
messages flow between R instances or other-language messaging components.
The key functions are:
aqSubscribe(aChannel) - subscribe to messages in a channel
aqWaitForData() - a blocking call that waits for data to arrive
aqPoll() - fetches all messages from the internal buffer (not the STOMP
server)
aqSend(channel, text) - send a message to a channel
This extension does buffer incoming data between consecutive aqPoll()
calls.
The core loop for using this should look like:
while(aqWaitForData()){
data = aqPoll()
}
Version 0.2 is beta. I have thoroughly tested it on Windows and on
Linux. I am looking for more testers and feedback.
Demo video at [2].
<buzz>I hope it opens up an easy venue to build distributed calculators
without the hassle of going through MPI or anything similar, in a truely
cross-platform approach, making R a viable citizen in highly
heterogenous message processing environments. </buzz>
Cheers,
Ulrich
References:
[1] http://stomp.github.com/implementations.html#STOMP_1_0_Servers
[2] http://www.youtube.com/watch?v=h1gLgJOEWW0
--
Ulrich Staudinger, ActiveQuant GmbH
P: +41 79 702 05 95
E:ustaudinger at activequant.com
http://www.activequant.com
Connect online:https://www.xing.com/profile/Ulrich_Staudinger
[[alternative HTML version deleted]]
_______________________________________________
R-packages mailing list
R-packages at r-project.org
https://stat.ethz.ch/mailman/listinfo/r-packages
More information about the R-help
mailing list