linkedin / spark-tfrecord

Read and write Tensorflow TFRecord data from Apache Spark.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

How to convert spark.ml.linalg.VectorUDT datatype to tfrecord?

mullerhai opened this issue · comments

HI :
I got the data after feature engineer with spark ,the dataframe column type is spark.ml.linalg.VectorUDT ,when I want to convert it to tfrecord meet error.

I don't think VectorUDT is supported. You will need to convert it to ArrayType(FloatType), assuming it's a vector of floats.

I don't think VectorUDT is supported. You will need to convert it to ArrayType(FloatType), assuming it's a vector of floats.

yeah, convert to ArrayType it is ok !