Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

predict returned empty results.

See original GitHub issue

predictions = fitted_model.transform(encoded)
predictions.show(1)
+--------------------+--------------+---------+
|            features|        labels|predicted|
+--------------------+--------------+---------+
|(784,[132,133,134...|(10,[1],[1.0])|       []|
+--------------------+--------------+---------+
only showing top 1 row

I got empty predicted, all the source code is here, https://github.com/huseinzol05/Pyspark-ML/blob/master/sparkflow/1.mnist-sparkflow-feedforward.ipynb.

And I use docker to spawn a spark environment, can check here, https://github.com/huseinzol05/Pyspark-ML/blob/master/Dockerfile-cluster

Again, thanks!

Issue Analytics

State:
Created 5 years ago
Comments:8 (6 by maintainers)

Top GitHub Comments

1reaction

dmmiller612commented, Jan 8, 2019

Added the prediction fix. This will have a breaking change for pyspark users using python 2 (We still support python 2, just a small behavior change). I will update sparkflow to 0.6 in pypi when I get home tonight.

0reactions

dmmiller612commented, Jan 9, 2019

Version 0.6.0 is deployed. I also included a Dockerfile with a Makefile to test the simple-dnn file.