Datatype conversion issue while requesting shapley values

[arunaryasomayajula]

H2O version, Operating System and Environment
3.34.0.8+ Pysparkling issue. Ticket #https://support.h2o.ai/a/tickets/109423 from Well Fargo

Actual behavior

Predicting from H2O3 model with shapley values produces error as below screenshot. This is due to a datatype conversion error and was fixed using the below. WF asking this to be part of the pysparkling package.

Expected behavior
No error should occur and the datatype conversion should happen within pysparkling.

Fix used by customer.

convert LongType columns in the DataFrame to DoubleType before the model.transform(dataFrame) .

See fix below:

for col_name in df.columns:
if df.schema[col_name].dataType == LongType():
df = df.withColumn(col_name, F.col(col_name).cast(DoubleType()))

Steps to reproduce
Steps to reproduce the behavior (with working code on a sample dataset, if possible):

1. Do this
2. Do that
3. Do something else
4. See error

Upload logs
If you can, please upload the H2O logs. More information on how to do that is available here, or you can use the h2o.downloadAllLogs() in R or the h2o.download_all_logs() function in Python.

Screenshots
If applicable, add screenshots to help explain your problem.

Additional context
Add any other context about the problem here.