Trying to fit a Linear Regression model in Spark I kept getting the error “`Column features must be of type org.apache.spark.ml.linalg.VectorUDT`”, I struggled looking for a solution, but finally found the key.