You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Upon a little experimentation, it seems the only thing preventing a big.matrix from the bigmemory package to be used as a Seed inside a DelayedArray is that dim() on a big.matrix returns a numeric and not an integer. This is caught by the check in extract_array() raising an error.
Would it be possible to relax this check to accept numeric as well as integer? Many on-disk matrix packages return numeric via dim() (bigmemory, filematrix, feather, fst, etc). It seems inconvenient to have to define a new Seed class for these packages just to reimplement the simple dim() function.
The text was updated successfully, but these errors were encountered:
Upon a little experimentation, it seems the only thing preventing a big.matrix from the bigmemory package to be used as a Seed inside a DelayedArray is that dim() on a big.matrix returns a numeric and not an integer. This is caught by the check in extract_array() raising an error.
Would it be possible to relax this check to accept numeric as well as integer? Many on-disk matrix packages return numeric via dim() (bigmemory, filematrix, feather, fst, etc). It seems inconvenient to have to define a new Seed class for these packages just to reimplement the simple dim() function.
The text was updated successfully, but these errors were encountered: