Renaming a column when opening a Parquet file in Pyspark / AWS Glue makes all data null

Question

I have a snappy.parquet file which I need to open as a DataFrame in spark, then upload to a database.
Two of the column names contain spaces (" ").
Usi