Closed yashwanthmadaka24 closed 5 years ago
the parameters given to a decision tree are "labelCol" which is of the type "double" and "featuresCol" which is a vector type.
JPMML-SparkML supports only a subset of Apache Spark ML types.
Specifically, the vector type is not supported. A vector column must be expanded into scalar columns, for example, by applying the VectorIndexer
transformation.
@yashwanthmadaka24 Hello! I know this thread is kind of old, but do you happen to remember how you solved this particular issue? I tried VectorIndexer but even after that it shows up as a vector. It is most likely that I did it wrongly, so I just want to know how you did it in particular.
@yashwanthmadaka24 Hello! I know this thread is kind of old, but do you happen to remember how you solved this particular issue? I tried VectorIndexer but even after that it shows up as a vector. It is most likely that I did it wrongly, so I just want to know how you did it in particular.
input col -> cast('double')
The above block of code throws the below error:
According to the documentation (https://spark.apache.org/docs/1.5.2/ml-decision-tree.html), the parameters given to a decision tree are "labelCol" which is of the type "double" and "featuresCol" which is a vector type. My trainingData also contains the exact same format. Is there any support for vectors?