Sadie has a Jython script which loads data and then uses slice1 to get separate components, and this fails with an index-out-of-bounds. This is an equivalent script:
xy= randn(0,2)
x= xy[:,0]
y= xy[:,1]
This should be supported because QUBE datasets allow for length(0) to be called when there is no zeroth record.
What makes this ticket-worthy is it's not clear what should happen when the data is a non-qube. The logic is the old behavior, but I'm sure there are many non-qubes which slice1 property when they have at least one record.
Sadie has a Jython script which loads data and then uses slice1 to get separate components, and this fails with an index-out-of-bounds. This is an equivalent script:
This should be supported because QUBE datasets allow for length(0) to be called when there is no zeroth record.
What makes this ticket-worthy is it's not clear what should happen when the data is a non-qube. The logic is the old behavior, but I'm sure there are many non-qubes which slice1 property when they have at least one record.