Bam files with no unmapped reads fails to sort

[ghost]

I have some exomes that I use for testing. One of them is from a mapped whole genome where I took a subset, reverted to fastq, and then removed unpaired reads. Looks like this resulted in there being no unmapped reads.

Looks like this code is where the error is coming from. Since the structure of unmapped reads is empty, it throws this error

Exception in thread "main" java.lang.UnsupportedOperationException: empty.max
	at scala.collection.TraversableOnce$class.max(TraversableOnce.scala:229)
	at scala.collection.AbstractTraversable.max(Traversable.scala:104)
	at org.bdgenomics.adam.rdd.read.AlignmentDataset$$anonfun$sortByReferencePositionAndIndex$1.apply(AlignmentDataset.scala:1036)
	at org.bdgenomics.adam.rdd.read.AlignmentDataset$$anonfun$sortByReferencePositionAndIndex$1.apply(AlignmentDataset.scala:1028)
	at scala.Option.fold(Option.scala:158)
	at org.apache.spark.rdd.Timer.time(Timer.scala:48)

Edit: Actually. I'm getting this error with simulated reads as well. Might be user error

[ghost]

Sorry. More noise on my end. Typo error with file path and reference. It looks like the library will not complain about a required file not existing.

[heuermh]

Are there conditions we should be checking for in ADAM and are not? We're relying on Hadoop-BAM and htsjdk upstream doing the right thing in a lot of BAM code paths, and that is not always a safe idea.