DataFrame instance in notebooks cannot be cast(verify=true) to itself

[koperagen]

This appears to be a regression, works fine in 1.0.0-Beta2 but fails in dev

%use dataframe
val test = DataFrame.readJson("test.json")
test

fun AnyFrame.benchmark() = castTo(test).select { benchmark and primaryMetric.score and primaryMetric.scoreConfidence  }

test.benchmark()

test.json

java.lang.IllegalArgumentException: Column has schema:
 primaryMetric:
	score: Float
	scoreError: Float
	scoreConfidence: List<Float>
	scorePercentiles:
		0.0: Float
		50.0: Float
		90.0: Float
		95.0: Float
		99.0: Float
		99.9: Float
		99.99: Float
		99.999: Float
		99.9999: Float
		100.0: Float
	scoreUnit: String
	rawData: List<List<Float>>

 that differs from target schema:
 primaryMetric:
	rawData: List<List<Float>>
	score: Float
	scoreConfidence: List<Float>
	scoreError: Float
	scorePercentiles:
		0.0: Float
		100.0: Float
		50.0: Float
		90.0: Float
		95.0: Float
		99.0: Float
		99.9: Float
		99.99: Float
		99.999: Float
		99.9999: Float
	scoreUnit: String

[Jolanrensen]

Maybe I missed it when changing schema comparisons?

It should be able to cast if it MATCHES, however it might check EQUALS now.