Hi,

Currently the DML algorithm scripts are written to the base directory of
the following artifacts:
  systemml-0.10.0-incubating-SNAPSHOT.jar
  systemml-0.10.0-incubating-SNAPSHOT-inmemory.jar
  systemml-0.10.0-incubating-SNAPSHOT-sources.jar
  systemml-0.10.0-incubating-SNAPSHOT-standalone.jar

This makes the base directory of these jars look similar to this:

$ tree -L 1 systemml-0.10.0-incubating-SNAPSHOT
systemml-0.10.0-incubating-SNAPSHOT
├── ALS-CG.dml
├── ALS-DS.dml
├── ALS_predict.dml
├── ALS_topk_predict.dml
├── Cox-predict.dml
├── Cox.dml
├── CsplineCG.dml
├── CsplineDS.dml
├── GLM-predict.dml
├── GLM.dml
├── KM.dml
├── Kmeans-predict.dml
├── Kmeans.dml
├── LinearRegCG.dml
├── LinearRegDS.dml
├── META-INF
├── MultiLogReg.dml
├── PCA.dml
├── StepGLM.dml
├── StepLinearRegDS.dml
├── Univar-Stats.dml
├── apply-transform.dml
├── bivar-stats.dml
├── com
├── decision-tree-predict.dml
├── decision-tree.dml
├── l2-svm-predict.dml
├── l2-svm.dml
├── m-svm-predict.dml
├── m-svm.dml
├── naive-bayes-predict.dml
├── naive-bayes.dml
├── obsolete
├── org
├── random-forest-predict.dml
├── random-forest.dml
├── stratstats.dml
└── transform.dml

Should these instead be written to scripts/algorithms as in the case of
some of the other artifacts so that the base directories of these jars look
like the following?

$ tree -L 1 systemml-0.10.0-incubating-SNAPSHOT
systemml-0.10.0-incubating-SNAPSHOT
├── META-INF
├── com
├── org
└── scripts

I created https://issues.apache.org/jira/browse/SYSTEMML-688 for this
issue. Perhaps this would be a good update for the 0.10.0 release? Any
thoughts?

Deron

Reply via email to