PySpark can't read properly from MongoDB

Question

I have written this simple script that connects to a MongoDB(3.6.2) server and read a collection named \'datascience.nypd\'. The dataset is quite big (~7milions of documents).