NOTE: If you attempt to refresh the schema of a parameterized dataset based on a set of files, only the schema for the first file is checked for changes. If changes are detected, the other files are must contain those changes as well. This can lead to changes being assumed or undetected in later files and potential data corruption in the flow.