How to ignore a column?

Legacy User

Hi all,

My table is a csv file with columns X1,X2,Y1,Y2, and I would like to fit separately Y1 = f(X1,X2) and then Y2 = f(X1,X2).

How do I tell the rapid miner to ignore the column Y2 when fitting the Y1?

Victor

steffen

Hello Victor

using (for example) FeatureNameFilter" switches Attributes off. If you want to perform both fittings within a single process, I further the suggest the usage of "IOMultiplier" and "IOSelector" (see manual for details)

something like this:

<operator name="Root" class="Process" expanded="yes">
    <operator name="IOMultiplier" class="IOMultiplier">
        <parameter key="io_object"	value="ExampleSet"/>
    </operator>
    <operator name="skip_Y1" class="FeatureNameFilter">
        <parameter key="skip_features_with_name"	value="Y1"/>
    </operator>
    <operator name="here_goes_Y2=f(X1,X2)" class="OperatorChain" expanded="yes">
    </operator>
    <operator name="IOSelector" class="IOSelector">
        <parameter key="io_object"	value="ExampleSet"/>
    </operator>
    <operator name="skip_Y2" class="FeatureNameFilter">
        <parameter key="skip_features_with_name"	value="Y2"/>
    </operator>
    <operator name="here_goes_Y1=f(X1,X2)" class="OperatorChain" expanded="yes">
    </operator>
</operator>

hope this was helpful

greetings

Steffen

Legacy User

Yes it was.

Could you also add a feature to the data import wizard: column type = {id, attribute, label,...} -- add "Ignored" to this list ?

For the newbies like me.

IngoRM

We will think about it. Actually, you could also use the operator InteractiveFeatureWeighting if you want to deselect features by a click - but this can sometimes be a lot of work 

Cheers,
Ingo