Hi Vedrin, 
 Thank you for your very nice card!

On Thu, 2009-12-31 at 14:15 +0200, Vedrin Jeliazkov wrote:
> Dear All,
> 
> I'm happy to announce that we have just reached a significant
> milestone in the last day of 2009 -- successful automated (continuous)
> testing of MLR model training. The workflow involves 3 OpenTox
> services (dataset, feature and model) running at two different
> locations (Plovdiv, Bulgaria and Athens, Greece). These services are
> part of two independent implementations of the OpenTox API v1.1 and
> they work together flawlessly. Credits go to Pantelis and Nina -- well
> done guys!
> 

Thanks Vedrin!

> Detailed statistics are available here:
> 
> http://ambit.uni-plovdiv.bg/cgi-bin/smokeping.cgi?target=NTUA.Models.TrainMLR
> 
> The configured timeout is 27 seconds. As you can see, in some cases
> the MLR model training either needs more time to succeed or simply
> fails (see below) and this results in some loss, being recorded by
> SmokePing. Obviously there's some room for debugging and further
> performance improvement. It might be worth mentioning that the
> involved dataset includes 522 records and 10 MLR models are trained
> every 5 minutes.
> 
> While performing some manual testing I've noticed that from time to
> time (only once in a while) the MLR model training fails with a
> message like the one below:
> 
> - - - - - - - - 8< - - - - - - - -
> 
> ambit:/# time curl -iv -X POST -d
> 'dataset_uri=http://ambit.uni-plovdiv.bg:8080/ambit2/dataset/6&target=http://ambit.uni-plovdiv.bg:8080/ambit2/feature/11954'
> http://opentox.ntua.gr:3000/algorithm/mlr
> * About to connect() to opentox.ntua.gr port 3000 (#0)
> *   Trying 147.102.82.32... connected
> * Connected to opentox.ntua.gr (147.102.82.32) port 3000 (#0)
> > POST /algorithm/mlr HTTP/1.1
> > User-Agent: curl/7.18.2 (x86_64-pc-linux-gnu) libcurl/7.18.2 OpenSSL/0.9.8g zlib/1.2.3.3 libidn/1.8 libssh2/0.18
> > Host: opentox.ntua.gr:3000
> > Accept: */*
> > Content-Length: 122
> > Content-Type: application/x-www-form-urlencoded
> >
> < HTTP/1.1 500 The server encountered an unexpected condition which
> prevented it from fulfilling the request
> HTTP/1.1 500 The server encountered an unexpected condition which
> prevented it from fulfilling the request
> < Content-Type: text/plain
> Content-Type: text/plain
> < Date: Thu, 31 Dec 2009 11:55:43 GMT
> Date: Thu, 31 Dec 2009 11:55:43 GMT
> < Accept-Ranges: bytes
> Accept-Ranges: bytes
> < Server: Noelios-Restlet/2.0m3
> Server: Noelios-Restlet/2.0m3
> < Connection: close
> Connection: close
> < Transfer-Encoding: chunked
> Transfer-Encoding: chunked
> 
> <
> Error Report.
> TimeStamp: Thu Dec 31 06:55:43 EST 2009
> 
> Error #1
> Exception Details: java.lang.IllegalArgumentException: Can't delete
> class attribute
> Explanation: Severe Error while trying to build an MLR model.
> For debugging reasons we provide a brief list of the exceptions:
> - weka.core.Instances.deleteAttributeAt(Unknown Source)
> - org.opentox.algorithm.dataprocessing.DataCleanUp.removeStringAtts(DataCleanUp.java:42)
> - org.opentox.algorithm.trainer.MlrTrainer.train(MlrTrainer.java:82)
> - org.opentox.resource.Algorithm.post(Algorithm.java:143)
> - org.restlet.resource.ServerResource.doHandle(ServerResource.java:340)
> 
> 
> * Closing connection #0
> 
> real    0m2.206s
> user    0m0.004s
> sys     0m0.004s
> 
> I hope this issue wouldn't prove too difficult to be resolved :-)

When my RDF parser detects an empty string or a value that should be
numeric but cannot be cast as such, it overrides it and considers of it
to be missing. I'll try to fix that - I just need one more check in a
critical step of the RDF parsing, so I think it wont be very difficult.
The message "Can't Handle class attribute" (because the class attribute
is of type String) appears because there is some non-numeric entry for
the specified target feature which characterizes the whole feature as
String. Many issues will be fixed if we establish datatype declarations
for features.


Best Regards,
Pantelis

> 
> Kind regards,
> Vedrin
> _______________________________________________
> Development mailing list
> Development at opentox.org
> http://www.opentox.org/mailman/listinfo/development
>