You can use a case class and rdd and then convert it to dataframe.
The common syntax to create a dataframe directly from a file is as shown below for your reference.
val df = spark.read.option("header","true").option(inferSchema,"true").csv("")
if you are relying on in-built schema of the csv file.
And If you don't then you can create a schema such as:
val schema = StructType(Array(StructField("AirportID", IntegerType, true)))