As far as I can see, transform method accepts only dataframe. Is there a way to somehow apply model to transform a single vector?
            Asked
            
        
        
            Active
            
        
            Viewed 27 times
        
    0
            
            
         
    
    
        ov7a
        
- 1,497
- 4
- 15
- 34
2 Answers
2
            
            
        There is not, or at least there is not if you ask about core Spark. Related feature request is tracked by SPARK-10413 and SPARK-10884 but none have seen much progress so far.
There exist external libraries which might address this problem, some mentioned in How to serve a Spark MLlib model?.
 
    
    
        user8972505
        
- 21
- 1
0
            
            
        Try to change the vector to a single row data Frame. And then run the model on it
 
    
    
        Debasish
        
- 113
- 1
- 9