Hi Marocco, Turns out the latest *released *version of the wso2-gpl[1] repository doesn't include the latest fixes related to pmml. What you could try is, to build the repositories manually as explained in[2] and use that p2-repo to install the pmml features to your CEP server.
I have tried out your svm pmml with the latest wso2-gpl pmml version and it works fine. [1] https://github.com/wso2-gpl/siddhi [2] https://docs.wso2.com/display/CEP410/Installing+WSO2+GPL+Features On Wed, Apr 20, 2016 at 6:09 PM, <[email protected]> wrote: > Hi Fazlan, > > I tried before writing this post but I always got the error: > > Different definition same as output stream definition :StreamDefinition > > So I this that the output of the predict is different but I don't know how > it has to be.. > > Have you ever try with a svm model? Can you test please my example? (It's > very important for me this question > > Thank you very much, > > Silvia > > Fazlan Nazeem <[email protected]> ha scritto: > > Hi Marocco, >> >> What you need is something like this. >> >> @Import('inputStream:1.0.0') >> define stream inputTest (field_0 double, field_1 double, field_2 double, >> field_3 double, field_4 double, field_5 double, field_6 double, field_7 >> double, field_8 double, field_9 double, field_10 double, field_11 double, >> field_12 double, field_13 double, field_14 double, field_15 double, >> field_16 double, field_17 double, field_18 double, field_19 double, >> field_20 double, field_21 double, field_22 double, field_23 double, >> field_24 double, field_25 double, field_26 double, field_27 double, >> field_28 double, field_29 double, field_30 double, field_31 double, >> field_32 double); >> >> @Import('outputStream:1.0.0') >> define stream outputTest (field_0 double, field_1 double, field_2 double, >> field_3 double, field_4 double, field_5 double, field_6 double, field_7 >> double, field_8 double, field_9 double, field_10 double, field_11 double, >> field_12 double, field_13 double, field_14 double, field_15 double, >> field_16 double, field_17 double, field_18 double, field_19 double, >> field_20 double, field_21 double, field_22 double, field_23 double, >> field_24 double, field_25 double, field_26 double, field_27 double, >> field_28 double, field_29 double, field_30 double, field_31 double, >> field_32 double, *target string*); >> >> >> from inputTest#pmml:predict('myfile.pmml') >> select * >> insert into outputTest; >> >> >> you have to create two streams (inputStream, outputStream) and create a >> logger publisher for outputStream. This way when an even is received to >> the >> outputStream, the event will be logged in the server console with the >> prediction value named as "target". >> >> >> >> On Tue, Apr 19, 2016 at 5:42 PM, <[email protected]> wrote: >> >> Thank you for the explanation.. >>> But I'm not able to write the correct shiddi query.. >>> I have this file pmml: >>> >>> <?xml version="1.0" encoding="UTF-8" standalone="yes"?> >>> <PMML xmlns="http://www.dmg.org/PMML-4_2";> >>> <Header description="linear SVM"> >>> <Application name="Apache Spark MLlib"/> >>> <Timestamp>2016-04-12T07:06:37</Timestamp> >>> </Header> >>> <DataDictionary numberOfFields="34"> >>> <DataField name="field_0" optype="continuous" dataType="double"/> >>> <DataField name="field_1" optype="continuous" dataType="double"/> >>> <DataField name="field_2" optype="continuous" dataType="double"/> >>> <DataField name="field_3" optype="continuous" dataType="double"/> >>> <DataField name="field_4" optype="continuous" dataType="double"/> >>> <DataField name="field_5" optype="continuous" dataType="double"/> >>> <DataField name="field_6" optype="continuous" dataType="double"/> >>> <DataField name="field_7" optype="continuous" dataType="double"/> >>> <DataField name="field_8" optype="continuous" dataType="double"/> >>> <DataField name="field_9" optype="continuous" dataType="double"/> >>> <DataField name="field_10" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_11" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_12" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_13" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_14" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_15" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_16" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_17" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_18" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_19" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_20" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_21" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_22" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_23" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_24" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_25" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_26" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_27" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_28" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_29" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_30" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_31" optype="continuous" >>> dataType="double"/> >>> <DataField name="field_32" optype="continuous" >>> dataType="double"/> >>> <DataField name="target" optype="categorical" dataType="string"/> >>> </DataDictionary> >>> <RegressionModel modelName="linear SVM" functionName="classification" >>> normalizationMethod="none"> >>> <MiningSchema> >>> <MiningField name="field_0" usageType="active"/> >>> <MiningField name="field_1" usageType="active"/> >>> <MiningField name="field_2" usageType="active"/> >>> <MiningField name="field_3" usageType="active"/> >>> <MiningField name="field_4" usageType="active"/> >>> <MiningField name="field_5" usageType="active"/> >>> <MiningField name="field_6" usageType="active"/> >>> <MiningField name="field_7" usageType="active"/> >>> <MiningField name="field_8" usageType="active"/> >>> <MiningField name="field_9" usageType="active"/> >>> <MiningField name="field_10" usageType="active"/> >>> <MiningField name="field_11" usageType="active"/> >>> <MiningField name="field_12" usageType="active"/> >>> <MiningField name="field_13" usageType="active"/> >>> <MiningField name="field_14" usageType="active"/> >>> <MiningField name="field_15" usageType="active"/> >>> <MiningField name="field_16" usageType="active"/> >>> <MiningField name="field_17" usageType="active"/> >>> <MiningField name="field_18" usageType="active"/> >>> <MiningField name="field_19" usageType="active"/> >>> <MiningField name="field_20" usageType="active"/> >>> <MiningField name="field_21" usageType="active"/> >>> <MiningField name="field_22" usageType="active"/> >>> <MiningField name="field_23" usageType="active"/> >>> <MiningField name="field_24" usageType="active"/> >>> <MiningField name="field_25" usageType="active"/> >>> <MiningField name="field_26" usageType="active"/> >>> <MiningField name="field_27" usageType="active"/> >>> <MiningField name="field_28" usageType="active"/> >>> <MiningField name="field_29" usageType="active"/> >>> <MiningField name="field_30" usageType="active"/> >>> <MiningField name="field_31" usageType="active"/> >>> <MiningField name="field_32" usageType="active"/> >>> <MiningField name="target" usageType="target"/> >>> </MiningSchema> >>> <RegressionTable intercept="0.0" targetCategory="1"> >>> <NumericPredictor name="field_0" >>> coefficient="-33.663867609108465"/> >>> <NumericPredictor name="field_1" >>> coefficient="-26.928869029283113"/> >>> <NumericPredictor name="field_2" >>> coefficient="-15.7589879271062"/> >>> <NumericPredictor name="field_3" >>> coefficient="4.801827945644439"/> >>> <NumericPredictor name="field_4" >>> coefficient="4.37913587227952"/> >>> <NumericPredictor name="field_5" >>> coefficient="5.017699542764569"/> >>> <NumericPredictor name="field_6" >>> coefficient="6.881223937543283"/> >>> <NumericPredictor name="field_7" >>> coefficient="7.436600728282555"/> >>> <NumericPredictor name="field_8" >>> coefficient="2.779684237173268"/> >>> <NumericPredictor name="field_9" >>> coefficient="-11.1563341379461"/> >>> <NumericPredictor name="field_10" >>> coefficient="-11.603667328060018"/> >>> <NumericPredictor name="field_11" >>> coefficient="-1.927205925226167"/> >>> <NumericPredictor name="field_12" >>> coefficient="-0.13042330215893916"/> >>> <NumericPredictor name="field_13" >>> coefficient="4.606266797295681"/> >>> <NumericPredictor name="field_14" >>> coefficient="4.198456448492548"/> >>> <NumericPredictor name="field_15" >>> coefficient="1.3866370557496586"/> >>> <NumericPredictor name="field_16" >>> coefficient="-1.6863769626309484"/> >>> <NumericPredictor name="field_17" >>> coefficient="-1.8279467318593428"/> >>> <NumericPredictor name="field_18" >>> coefficient="0.1893987855143311"/> >>> <NumericPredictor name="field_19" >>> coefficient="12.645268212559404"/> >>> <NumericPredictor name="field_20" >>> coefficient="14.115013455939593"/> >>> <NumericPredictor name="field_21" >>> coefficient="13.544000807391123"/> >>> <NumericPredictor name="field_22" >>> coefficient="6.0112812628907015"/> >>> <NumericPredictor name="field_23" >>> coefficient="4.03765113349721"/> >>> <NumericPredictor name="field_24" >>> coefficient="4.145995153508418"/> >>> <NumericPredictor name="field_25" >>> coefficient="2.08758768566234"/> >>> <NumericPredictor name="field_26" >>> coefficient="1.1271455002504913"/> >>> <NumericPredictor name="field_27" >>> coefficient="-0.28071996147406497"/> >>> <NumericPredictor name="field_28" >>> coefficient="7.761062281510407"/> >>> <NumericPredictor name="field_29" >>> coefficient="0.25888510113432384"/> >>> <NumericPredictor name="field_30" >>> coefficient="-8.756880183958602"/> >>> <NumericPredictor name="field_31" >>> coefficient="-4.332351913516022"/> >>> <NumericPredictor name="field_32" >>> coefficient="2.0548737734231493"/> >>> </RegressionTable> >>> <RegressionTable intercept="0.0" targetCategory="0"/> >>> </RegressionModel> >>> </PMML> >>> >>> so I have to write something like: >>> >>> @Import('fonometro:1.0.0') >>> define stream inputTest (field_0 double, field_1 double, field_2 double, >>> field_3 double, field_4 double, field_5 double, field_6 double, field_7 >>> double, field_8 double, field_9 double, field_10 double, field_11 double, >>> field_12 double, field_13 double, field_14 double, field_15 double, >>> field_16 double, field_17 double, field_18 double, field_19 double, >>> field_20 double, field_21 double, field_22 double, field_23 double, >>> field_24 double, field_25 double, field_26 double, field_27 double, >>> field_28 double, field_29 double, field_30 double, field_31 double, >>> field_32 double); >>> >>> from inputTest#pmml:predict('samples/myfile.pmml') >>> select * >>> insert into ???? >>> >>> or >>> >>> @Import('fonometro:1.0.0') >>> define stream inputTest (field_0 double, field_1 double, field_2 double, >>> field_3 double, field_4 double, field_5 double, field_6 double, field_7 >>> double, field_8 double, field_9 double, field_10 double, field_11 double, >>> field_12 double, field_13 double, field_14 double, field_15 double, >>> field_16 double, field_17 double, field_18 double, field_19 double, >>> field_20 double, field_21 double, field_22 double, field_23 double, >>> field_24 double, field_25 double, field_26 double, field_27 double, >>> field_28 double, field_29 double, field_30 double, field_31 double, >>> field_32 double); >>> >>> from inputTest#pmml:predict('myfile.pmml',field_0, field_1, field_2, >>> field_3, field_4, field_5, field_6, field_7, field_8, field_9, field_10, >>> field_11, field_12, field_13, field_14, field_15, field_16, field_17, >>> field_18, field_19, field_20, field_21, field_22, field_23, field_24, >>> field_25, field_26, field_27, field_28, field_29, field_30, field_31, >>> field_32) >>> select * >>> insert into ???? >>> >>> everything I put instead of ???? raise an error saying that I'm >>> redefining >>> the output stream.. >>> >>> Can you help me? I need the "target" field? >>> Thank you, >>> >>> Silvia >>> >>> Quoting Fazlan Nazeem <[email protected]>: >>> >>> Hi Marocco, >>> >>>> >>>> The reason behind the decision tree PMML model having an output tag and >>>> SVM >>>> not having an output tag solely depends on what library you have used to >>>> generate the PMML file. The PMML specification[1] does not specify that >>>> every PMML model should include an output tag. Please refer to the line >>>> 7 >>>> of the XSD file attached herewith which specifies that output tag >>>> "minOccurs="0", which implies that a PMML file may contain a minimum of >>>> 0 >>>> output fields. For example you may try out WSO2 Machine Learner's[2] >>>> PMML >>>> generation functionality which does not include output tags for most of >>>> its >>>> PMML models. >>>> >>>> Therefore two specific libraries for PMML can generate two slightly >>>> different PMML files for the same model depending on how they have >>>> implemented the PMML generation functionality. The PMML readers should >>>> be >>>> able to interpret these different cases and predict the output for a new >>>> input stream. >>>> >>>> With CEP 4.0.0, you may experience an issue when there is no output tag >>>> present in the PMML file, the PMML extension might not work. We have >>>> identified this issue[3] and fixed it with the 4.1.0 release. >>>> >>>> >>>> [1] http://dmg.org/pmml/v4-1/GeneralStructure.html >>>> [2] http://wso2.com/products/machine-learner/ >>>> [3] https://wso2.org/jira/browse/CEP-1438 >>>> >>>> On Fri, Apr 15, 2016 at 8:12 PM, <[email protected]> wrote: >>>> >>>> >>>> Hi! >>>>> >>>>> I installed the cep version 4, becouse I need to test the pmml >>>>> functionalities. I have already executed this example: >>>>> >>>>> >>>>> https://docs.wso2.com/display/CEP410/Sample+0301+-+Network+Intruder+Detection+with+PMML+Extension+Predictions >>>>> which is based on a decision-tree and it works. I need to understand >>>>> for >>>>> a >>>>> SVM (support vector machine) PMML, which is the output stream I have to >>>>> create on CEP. The difference between the two pmml is that >>>>> decision-tree >>>>> specifies Output tag, so the outputstream is equal to the inputstrams >>>>> plus >>>>> the output tag, in svm instead there is no output tag. >>>>> >>>>> Thank you for help. >>>>> Silvia >>>>> >>>>> _______________________________________________ >>>>> Dev mailing list >>>>> [email protected] >>>>> http://wso2.org/cgi-bin/mailman/listinfo/dev >>>>> >>>>> >>>>> >>>> >>>> -- >>>> Thanks & Regards, >>>> >>>> Fazlan Nazeem >>>> >>>> *Software Engineer* >>>> >>>> *WSO2 Inc* >>>> Mobile : +94772338839 >>>> <%2B94%20%280%29%20773%20451194> >>>> [email protected] >>>> >>>> >>> >>> >>> >>> >> >> -- >> Thanks & Regards, >> >> Fazlan Nazeem >> >> *Software Engineer* >> >> *WSO2 Inc* >> Mobile : +94772338839 >> <%2B94%20%280%29%20773%20451194> >> [email protected] >> > > > -- Thanks & Regards, Fazlan Nazeem *Software Engineer* *WSO2 Inc* Mobile : +94772338839 <%2B94%20%280%29%20773%20451194> [email protected]
_______________________________________________ Dev mailing list [email protected] http://wso2.org/cgi-bin/mailman/listinfo/dev
