ArrayIndexOutOfBoundsException

marvin_souza

Hi all.
I am using the RapidMiner to do Task Inference in my Master tesis. After some tests with different data, occured a strange error.
First I will describe my scenario.
XML:

<operator name="Root" class="Process" expanded="yes">
    <operator name="DatabaseExampleSource" class="DatabaseExampleSource">
        <parameter key="database_system"	value="PostgreSQL"/>
        <parameter key="database_url"	value="jdbc:postgresql://parara:5432/exehda"/>
        <parameter key="username"	value="marcos"/>
        <parameter key="password"	value="vBwyakUxtRE="/>
        <parameter key="query"	value="SELECT  t.name as task,h.sector, h.time_frame, h.main_device, sec.info, sec.type  FROM  exehda_history h left join exehda_history_sec_ctx sec on sec.id_history=h.id  join exehda_task t on h.id_task = t.id where h.id_user = 1"/>
        <parameter key="label_attribute"	value="task"/>
        <parameter key="datamanagement"	value="double_sparse_array"/>
    </operator>
    <operator name="DecisionTree" class="DecisionTree">
        <parameter key="criterion"	value="information_gain"/>
        <parameter key="minimal_size_for_split"	value="5"/>
        <parameter key="minimal_leaf_size"	value="3"/>
        <parameter key="minimal_gain"	value="0.2"/>
        <parameter key="number_of_prepruning_alternatives"	value="2"/>
        <parameter key="no_pre_pruning"	value="true"/>
        <parameter key="no_pruning"	value="true"/>
    </operator>
</operator>

My data:

AtendimentoPaciente	clinica	43	desktop	(null)	(null)
AtendimentoPaciente	clinica	42	desktop	(null)	(null)
AtendimentoPaciente	clinica	42	desktop	(null)	(null)
AtendimentoPaciente	clinica	41	desktop	(null)	(null)
AtendimentoPaciente	clinica	41	desktop	(null)	(null)
AtendimentoPaciente	clinica	40	desktop	(null)	(null)
AtendimentoPaciente	clinica	40	desktop	(null)	(null)
Cirurgia	        clinica	45	desktop	(null)	(null)

When I run this proccess, I got the error:

java.lang.ArrayIndexOutOfBoundsException: 0
    at com.rapidminer.operator.learner.tree.InfoGainCriterion.getBenefit(InfoGainCriterion.java:66)
    at com.rapidminer.operator.learner.tree.InfoGainCriterion.getNominalBenefit(InfoGainCriterion.java:56)
    at com.rapidminer.operator.learner.tree.TreeBuilder.calculateBenefit(TreeBuilder.java:108)
    at com.rapidminer.operator.learner.tree.TreeBuilder.calculateAllBenefits(TreeBuilder.java:134)
    at com.rapidminer.operator.learner.tree.TreeBuilder.buildTree(TreeBuilder.java:157)
    at com.rapidminer.operator.learner.tree.TreeBuilder.learnTree(TreeBuilder.java:93)
    at com.rapidminer.operator.learner.tree.AbstractTreeLearner.learn(AbstractTreeLearner.java:109)
    at com.rapidminer.operator.learner.AbstractLearner.apply(AbstractLearner.java:88)
    at com.rapidminer.operator.Operator.apply(Operator.java:671)
    at com.rapidminer.operator.OperatorChain.apply(OperatorChain.java:424)
    at com.rapidminer.operator.Operator.apply(Operator.java:671)
    at com.rapidminer.Process.run(Process.java:735)
    at com.rapidminer.Process.run(Process.java:704)
    at com.rapidminer.Process.run(Process.java:694)
    at com.rapidminer.gui.ProcessThread.run(ProcessThread.java:59)

If the data is generated randomically, this don't happen. Some suggestion ?

marvin_souza

By some obscure reason, the RM don't handles the null values.
When I remove the columns info and type, everything works fine.

steffen

Hello marvin

I assume that the data is loaded as "(null)", not as missing values. Correct ?
Another Question: Do the mentioned columns have each only one unique value ?

regards,

Steffen