Closed xrotwang closed 1 year ago
This is dealt with in most examples. The long-to-wide issue cannot be addressed generically, because there is no general rule on datasets having just one datapoint per language/parameter (many have not e.g. PHOIBLE, APiCS, many Wordlists).
See https://github.com/cldf/pycldf/issues/82#issuecomment-488645237