Loading new data

[tercer]

Need new complete workflow for running PSI more seamlessly on new dataset.

[tercer]

Currently we use both the Dataverse .xml metadata (for reading row numbers and variable names) as well as the JSON ravens_metadata product for reading typing information (numchar, nature, binary). So these currently have to be created for new datasets.
Could obviously get names and row numbers also from ravens metadata.

[tercer]

Utility code 'xmlutility.R' in '/rook' will create skeleton .xml document with necessary fields in style of Dataverse schema.
http://metadata.2ravens.org/ works for JSON file.