There seems to be no way to define data classes where the data class encoder produces a Spark schema with fields of type Decimal(38, 0). The natural approach would be to define a data class with a field of type BigInteger, but this is unsupported by the data class encoder.

This can be seen by the following code

data class A(val value: BigInteger) fun main() = withSpark { val ds = dsOf(1, 2) val df = ds.`as`<A>() println(df.schema()) }

which throws
java.lang.IllegalArgumentException: java.math.BigInteger is unsupported.