[SOLVED] Issue: how to implemet a dynamic filter on Rapidminer

jseeker

Dear community members,
I'm using RapidMiner and RapidAnalytics for analyse a dataset, downloaded from http://archive.ics.uci.edu/ml/machine-learning-databases/00257/Data_User_Modeling_Dataset_Hamdi%20Tolga%20KAHRAMAN.xls
I want to filter my dataset by UNS column.
In rapidMiner I already implemented a process like this and I export this process on RapidAnalytics.
How we can see the last block is a filter example with attribute_value_filter set on UNS with regular expression value.
Now i'm going to implement a dynamic filter, but i dont know how do that.
I have to filter data by label : Very Low, Low, Middle and High.
My final result should be a dynamic process on RapidAnalytics that allows to see four data view.
Anyone could help me ?

Thank in advance.

bye

MariusHelf

Hi,

how do you access RapidAnalytics? Are you using a webservice? In that case you could make use of macros.

In RapidMiner define a macro called "filterValue" in the Process Context (you find it next to the Parameters tab in the upper right).
Then define your filter as

UNS = .*%{filterValue}.*

This allows you to provide the filterValue as a macro value in RapidMiner. This macro value can also be exported to the webservice, and you can call it by something like http://my.rapidminer.server.com/myWebservice?filterValue=High

Best regards,
Marius

jseeker

Hi Marius,
thank you so much for the answer.
I access to rapidAnalytics with firefox browser and i use a webservice.
Added the macro in rapidMiner, image attached

Now , I would export this macro into webservice, but when i set up the service i don't find anything.

Right now, i tryed to insert the macro in url query parameter like this

My result on service test is

I typed the low value but the filter doesn't work
where is my mistake?

Thank you

Regards.

MariusHelf

Hi,

you are using the Set Macro operator, but to make the filter work you have to use the process context as described above.

Best regards,
Marius

jseeker

Hi,
unfortunately i've no result in my table.
It's becoming a nightmare.
I'm going to write what i did. So, i restored my previous process context in rapidMiner.
I'm using a filter example with :
- condition_class -> attribute_value_filter
- parameter string -> UNS = .*%{filterValue}.*
I even exported the process in rapidAnalytic such as service and added the same rapidMiner's macro.

In Service Settings
Parameter binding :
- URL query parameter -> filterValue
- Target (macro/operator parameter) -> UNS = .*%{filterValue}.*

when I tried to test the service, it request me the parameter, i write for example Low and my result always is empty table

Why ?

I must finish this project as soon as possibile.
And you Marius are my hope.

Thank for the patience
Regards.

Marco_Boeck

Hi,

in the service settings on the server, URL-Query-Parameter and Target must be identical, i.e. "filterValue" in your case.

Testprocess:

<?xml version="1.0" encoding="UTF-8" standalone="no"?>
<process version="5.3.014">
  <context>
    <input/>
    <output/>
    <macros>
      <macro>
        <key>filterValue</key>
        <value/>
      </macro>
    </macros>
  </context>
  <operator activated="true" class="process" compatibility="5.3.014" expanded="true" name="Process">
    <process expanded="true">
      <operator activated="true" class="generate_nominal_data" compatibility="5.3.014" expanded="true" height="60" name="Generate Nominal Data" width="90" x="45" y="30"/>
      <operator activated="true" class="filter_examples" compatibility="5.3.014" expanded="true" height="76" name="Filter Examples" width="90" x="179" y="30">
        <parameter key="parameter_string" value="att1 = .*%{filterValue}.*"/>
        <parameter key="condition_class" value="attribute_value_filter"/>
        <list key="filters_list"/>
      </operator>
      <connect from_op="Generate Nominal Data" from_port="output" to_op="Filter Examples" to_port="example set input"/>
      <connect from_op="Filter Examples" from_port="example set output" to_port="result 1"/>
      <portSpacing port="source_input 1" spacing="0"/>
      <portSpacing port="sink_result 1" spacing="0"/>
      <portSpacing port="sink_result 2" spacing="0"/>
    </process>
  </operator>
</process>

These two must be identical and match the macro value in the process context and the filter operator


This is an example of a test with the filter set to "value0".


Regards,
Marco

jseeker

Hi,
Finally i solved my problem
Thank you so much.
I'll open very soon another post for how to predict the level of user's knowledge.
I'm using the same dataset, i have five attributes and one UNS label, that says me  the level of user belongs.
I'm very thankful for your precious help.

Regards.