use of com.alibaba.alink.pipeline.clustering.Lda in project Alink by alibaba.
the class PipelineModelTest method getPipeline.
protected Pipeline getPipeline() {
// model mapper
QuantileDiscretizer quantileDiscretizer = new QuantileDiscretizer().setNumBuckets(2).setSelectedCols("sepal_length");
// SISO mapper
Binarizer binarizer = new Binarizer().setSelectedCol("petal_width").setOutputCol("bina").setReservedCols("sepal_length", "petal_width", "petal_length", "category").setThreshold(1.);
// MISO Mapper
VectorAssembler assembler = new VectorAssembler().setSelectedCols("sepal_length", "petal_width").setOutputCol("assem").setReservedCols("sepal_length", "petal_width", "petal_length", "category");
// Lda
Lda lda = new Lda().setPredictionCol("lda_pred").setPredictionDetailCol("lda_pred_detail").setSelectedCol("category").setTopicNum(2).setRandomSeed(0);
return new Pipeline().add(binarizer).add(assembler).add(quantileDiscretizer).add(lda);
}
use of com.alibaba.alink.pipeline.clustering.Lda in project Alink by alibaba.
the class LocalPredictorTest method getPipeline.
protected Pipeline getPipeline() {
// model mapper
QuantileDiscretizer quantileDiscretizer = new QuantileDiscretizer().setNumBuckets(2).setSelectedCols("sepal_length");
// SISO mapper
Binarizer binarizer = new Binarizer().setSelectedCol("petal_width").setOutputCol("bina").setReservedCols("sepal_length", "petal_width", "petal_length", "category").setThreshold(1.);
// MISO Mapper
VectorAssembler assembler = new VectorAssembler().setSelectedCols("sepal_length", "petal_width").setOutputCol("assem").setReservedCols("sepal_length", "petal_width", "petal_length", "category");
// Lda
Lda lda = new Lda().setPredictionCol("lda_pred").setPredictionDetailCol("lda_pred_detail").setSelectedCol("category").setTopicNum(2).setRandomSeed(0);
Select select = new Select().setClause("cast(sepal_length as double) as sepal_length, " + "cast(petal_width as double) as petal_width, " + "cast(petal_length as double) as petal_length, " + "category");
// Glm
GeneralizedLinearRegression glm = new GeneralizedLinearRegression().setFeatureCols("sepal_length", "petal_width").setLabelCol("petal_length").setPredictionCol("glm_pred");
return new Pipeline().add(binarizer).add(assembler).add(quantileDiscretizer).add(glm);
}
Aggregations