odps/static/algorithms/preprocess.xml

<?xml version='1.0' encoding='UTF-8'?>
<algorithms baseClass="BaseProcessAlgorithm">
    <algorithm codeName="AppendID">
        <public>false</public>
        <reloadFields>false</reloadFields>
        <fieldChangable>false</fieldChangable>
        <params>
            <param name="inputTableName">
                <exporter>get_input_table_name</exporter>
                <inputName>input</inputName>
            </param>
            <param name="inputPartitions">
                <exporter>get_input_partitions</exporter>
                <inputName>input</inputName>
            </param>
            <param name="outputTableName">
                <exporter>get_output_table_name</exporter>
                <outputName>output</outputName>
            </param>
            <param name="outputPartition">
                <exporter>get_output_table_partitions</exporter>
                <outputName>output</outputName>
            </param>
            <param name="IDColName">
                <alias>idCol</alias>
                <value>append_id</value>
            </param>
            <param name="selectedColNames">
                <exporter>$package_root.preprocess._customize.get_append_id_selected_cols</exporter>
                <alias>selectedCols</alias>
                <inputName>input</inputName>
            </param>
        </params>
        <ports>
            <port name="input">
                <ioType>INPUT</ioType>
                <sequence>1</sequence>
                <type>DATA</type>
            </port>
            <port name="output">
                <ioType>OUTPUT</ioType>
                <sequence>1</sequence>
                <type>DATA</type>
                <schema>
                    <schema>{IDColName}: bigint:, {selectedColNames}</schema>
                </schema>
            </port>
        </ports>
        <metas>
            <meta name="xflowName" value="AppendID"/>
            <meta name="xflowProjectName" value="algo_public"/>
        </metas>
    </algorithm>
    <algorithm codeName="Split">
        <public>false</public>
        <reloadFields>false</reloadFields>
        <fieldChangable>false</fieldChangable>
        <params>
            <param name="inputTableName">
                <exporter>get_input_table_name</exporter>
                <inputName>input</inputName>
            </param>
            <param name="inputTablePartitions">
                <exporter>get_input_partitions</exporter>
                <inputName>input</inputName>
            </param>
            <param name="output1TableName">
                <exporter>get_output_table_name</exporter>
                <outputName>output1</outputName>
            </param>
            <param name="output1TablePartition">
                <exporter>get_output_table_partitions</exporter>
                <outputName>output1</outputName>
            </param>
            <param name="output2TableName">
                <exporter>get_output_table_name</exporter>
                <outputName>output2</outputName>
            </param>
            <param name="output2TablePartition">
                <exporter>get_output_table_partitions</exporter>
                <outputName>output2</outputName>
            </param>
            <param name="fraction" />
            <param name="randomSeed" />
        </params>
        <ports>
            <port name="input">
                <ioType>INPUT</ioType>
                <sequence>1</sequence>
                <type>DATA</type>
            </port>
            <port name="output1">
                <ioType>OUTPUT</ioType>
                <sequence>1</sequence>
                <type>DATA</type>
            </port>
            <port name="output2">
                <ioType>OUTPUT</ioType>
                <sequence>2</sequence>
                <type>DATA</type>
            </port>
        </ports>
        <metas>
            <meta name="xflowName" value="Split"/>
            <meta name="xflowProjectName" value="algo_public"/>
        </metas>
    </algorithm>
    <algorithm codeName="RandomSample">
        <public>false</public>
        <reloadFields>false</reloadFields>
        <fieldChangable>false</fieldChangable>
        <params>
            <param name="inputTableName">
                <exporter>get_input_table_name</exporter>
                <inputName>input</inputName>
            </param>
            <param name="inputTablePartitions">
                <exporter>get_input_partitions</exporter>
                <inputName>input</inputName>
            </param>
            <param name="outputTableName">
                <exporter>get_output_table_name</exporter>
                <outputName>output</outputName>
            </param>
            <param name="sampleSize" />
            <param name="sampleRatio" />
            <param name="replace" />
            <param name="randomSeed" />
        </params>
        <ports>
            <port name="input">
                <ioType>INPUT</ioType>
                <sequence>1</sequence>
                <type>DATA</type>
            </port>
            <port name="output">
                <ioType>OUTPUT</ioType>
                <sequence>1</sequence>
                <type>DATA</type>
            </port>
        </ports>
        <metas>
            <meta name="xflowName" value="RandomSample"/>
            <meta name="xflowProjectName" value="algo_public"/>
        </metas>
    </algorithm>
    <algorithm codeName="WeightedSample">
        <public>false</public>
        <reloadFields>false</reloadFields>
        <fieldChangable>false</fieldChangable>
        <params>
            <param name="inputTableName">
                <exporter>get_input_table_name</exporter>
                <inputName>input</inputName>
            </param>
            <param name="inputTablePartitions">
                <exporter>get_input_partitions</exporter>
                <inputName>input</inputName>
            </param>
            <param name="outputTableName">
                <exporter>get_output_table_name</exporter>
                <outputName>output</outputName>
            </param>
            <param name="sampleSize" />
            <param name="sampleRatio" />
            <param name="probCol" />
            <param name="replace" />
            <param name="randomSeed" />
        </params>
        <ports>
            <port name="input">
                <ioType>INPUT</ioType>
                <sequence>1</sequence>
                <type>DATA</type>
            </port>
            <port name="output">
                <ioType>OUTPUT</ioType>
                <sequence>1</sequence>
                <type>DATA</type>
            </port>
        </ports>
        <metas>
            <meta name="xflowName" value="WeightedSample"/>
            <meta name="xflowProjectName" value="algo_public"/>
        </metas>
    </algorithm>
    <algorithm codeName="StratifiedSample">
        <public>false</public>
        <reloadFields>false</reloadFields>
        <fieldChangable>false</fieldChangable>
        <params>
            <param name="inputTableName">
                <exporter>get_input_table_name</exporter>
                <inputName>input</inputName>
            </param>
            <param name="inputTablePartitions">
                <exporter>get_input_partitions</exporter>
                <inputName>input</inputName>
            </param>
            <param name="outputTableName">
                <exporter>get_output_table_name</exporter>
                <outputName>output</outputName>
            </param>
            <param name="sampleSize" />
            <param name="sampleRatio" />
            <param name="strataColName" />
            <param name="randomSeed" />
        </params>
        <ports>
            <port name="input">
                <ioType>INPUT</ioType>
                <sequence>1</sequence>
                <type>DATA</type>
            </port>
            <port name="output">
                <ioType>OUTPUT</ioType>
                <sequence>1</sequence>
                <type>DATA</type>
            </port>
        </ports>
        <metas>
            <meta name="xflowName" value="StratifiedSample"/>
            <meta name="xflowProjectName" value="algo_public"/>
        </metas>
    </algorithm>
    <algorithm codeName="binning">
        <enabled>false</enabled>
        <reloadFields>false</reloadFields>
        <fieldChangable>false</fieldChangable>
        <params>
            <param name="inputTableName">
                <exporter>get_input_table_name</exporter>
                <inputName>feature</inputName>
            </param>
            <param name="inputPartitions">
                <exporter>get_input_partitions</exporter>
                <inputName>feature</inputName>
            </param>
            <param name="outputTableName">
                <exporter>get_output_table_name</exporter>
                <outputName>output</outputName>
            </param>
            <param name="selectedColNames">
                <alias>cols</alias>
                <exporter>get_feature_columns</exporter>
                <inputName>feature</inputName>
            </param>
            <param name="labelColName">
                <alias>labelCol</alias>
                <exporter>get_label_column</exporter>
                <inputName>feature</inputName>
            </param>
            <param name="nDivide">
                <value>10</value>
            </param>
            <param name="isLeftOpen"/>
            <param name="stringThreshold">
                <value>-1</value>
            </param>
            <param name="positiveLabel">
                <value>1</value>
            </param>
            <param name="inputBinTableName">
                <exporter>get_input_table_name</exporter>
                <inputName>bins</inputName>
            </param>
            <param name="binningMethod"/>
        </params>
        <ports>
            <port name="feature">
                <ioType>INPUT</ioType>
                <sequence>1</sequence>
                <type>DATA</type>
            </port>
            <port name="bins">
                <ioType>INPUT</ioType>
                <sequence>2</sequence>
                <type>DATA</type>
            </port>
            <port name="output">
                <ioType>OUTPUT</ioType>
                <sequence>1</sequence>
                <type>DATA</type>
                <schema>
                    <schema>feature: string, json: string</schema>
                </schema>
            </port>
        </ports>
        <metas>
            <meta name="xflowName" value="binning"/>
            <meta name="xflowProjectName" value="algo_public"/>
        </metas>
    </algorithm>
    <algorithm codeName="binning_training">
        <baseClass>BaseTrainingAlgorithm</baseClass>
        <reloadFields>false</reloadFields>
        <fieldChangable>false</fieldChangable>
        <enabled>false</enabled>
        <params>
            <param name="featureColNames">
                <alias>featureCols</alias>
                <exporter>get_feature_columns</exporter>
                <inputName>feature</inputName>
            </param>
            <param name="labelColName">
                <alias>labelCol</alias>
                <exporter>get_label_column</exporter>
                <inputName>feature</inputName>
            </param>
            <param name="optimization">
                <value>barrier_method</value>
            </param>
            <param name="loss">
                <value>logistic_regression</value>
            </param>
            <param name="iterations">
                <value>100</value>
            </param>
            <param name="l1Weight">
                <value>0</value>
            </param>
            <param name="l2Weight">
                <value>0</value>
            </param>
            <param name="m">
                <value>10</value>
            </param>
            <param name="convergenceTolerance">
                <value>0.000001</value>
            </param>
            <param name="positiveLabel">
                <value>1</value>
            </param>
            <param name="inputTableName" required="true">
                <exporter>get_input_table_name</exporter>
                <inputName>feature</inputName>
            </param>
            <param name="inputTablePartitions">
                <exporter>get_input_partitions</exporter>
                <inputName>feature</inputName>
            </param>
            <param name="outputTableName">
                <exporter>get_output_model_table_name(table_name=model)</exporter>
                <outputName>output</outputName>
            </param>
            <param name="inputBinTableName">
                <exporter>get_input_table_name</exporter>
                <inputName>bins</inputName>
            </param>
        </params>
        <ports>
            <port name="feature">
                <ioType>INPUT</ioType>
                <sequence>1</sequence>
                <type>DATA</type>
            </port>
            <port name="bins">
                <ioType>INPUT</ioType>
                <sequence>2</sequence>
                <type>DATA</type>
            </port>
            <port name="output">
                <ioType>OUTPUT</ioType>
                <sequence>1</sequence>
                <type>MODEL</type>
                <model>
                    <type>TablesModel</type>
                    <schemas>
                        <schema name="model">
                            <schema>feaname: string, binid: bigint, bin: string, constraint: string, weight: double,
                                scaled_weight: bigint, woe: double, constribution: double, total: bigint, positive:
                                bigint, negative: bigint, percetage_pos: double, percetage_neg: double
                            </schema>
                        </schema>
                        <schema name="bins">
                            <copyInput>bins</copyInput>
                            <directCopy>bins</directCopy>
                        </schema>
                    </schemas>
                </model>
            </port>
        </ports>
        <metas>
            <meta name="xflowName" value="linear_model"/>
            <meta name="xflowProjectName" value="algo_public"/>
            <meta name="predictor" value="binning_predict"/>
        </metas>
    </algorithm>
    <algorithm codeName="binning_predict">
        <public>false</public>
        <reloadFields>false</reloadFields>
        <enabled>false</enabled>
        <params>
            <param name="inputFeatureTableName">
                <exporter>get_input_table_name</exporter>
                <inputName>feature</inputName>
            </param>
            <param name="inputFeatureTablePartitions">
                <exporter>get_input_partitions</exporter>
                <inputName>feature</inputName>
            </param>
            <param name="inputBinTableName">
                <exporter>get_input_model_table_name(table_name=bins)</exporter>
                <inputName>model</inputName>
            </param>
            <param name="outputTableName">
                <exporter>get_output_table_name</exporter>
                <outputName>output</outputName>
            </param>
            <param name="inputModelTableName">
                <exporter>get_input_model_table_name(table_name=model)</exporter>
                <inputName>model</inputName>
            </param>
            <param name="featureColNames">
                <alias>featureCols</alias>
                <exporter>get_feature_columns</exporter>
                <inputName>feature</inputName>
            </param>
            <param name="metaColNames">
                <exporter>get_original_columns</exporter>
                <inputName>feature</inputName>
            </param>
        </params>
        <ports>
            <port name="model">
                <ioType>INPUT</ioType>
                <sequence>1</sequence>
                <type>MODEL</type>
            </port>
            <port name="feature">
                <ioType>INPUT</ioType>
                <sequence>2</sequence>
                <type>DATA</type>
            </port>
            <port name="output">
                <ioType>OUTPUT</ioType>
                <sequence>1</sequence>
                <type>DATA</type>
                <schema>
                    <schema>{metaColNames}, prediction_score: double, prediction_prob: double, prediction_detail: string</schema>
                </schema>
            </port>
        </ports>
        <metas>
            <meta name="xflowName" value="lm_predict"/>
            <meta name="xflowProjectName" value="algo_public"/>
        </metas>
    </algorithm>
</algorithms>