How to create a schema dynamically in pyspark where the data type is also decided dynamically

Question

I tried to build the schema programmatically but I am not able to get a fully workable solution. Any help here will be much appreciated.

def dynamic_schema_ge