Setting up column filtering --------------------------- The next step is to enable filtering on a per-column basis. Indicia has not automatically done this for you since it does not know the type of data in each field yet so it does not know how to write the SQL to filter the records. We can do this using the ``datatype`` attribute of a column definition, which can be set to date, text, integer or float. Try this by adding an attribute ``datatype="text"`` to the definition of the Species column, save the report file and reload the report in the warehouse. You should have an extra box at the top of the Species column so try typing the first few characters of a species name followed by * then tabbing out of the box to check it works. Next step, of course, is to apply the datatype attribute to the other column definitions. Your XML document should now look like: .. code-block:: xml select #columns# from cache_occurrences_functional o join cache_samples_nonfunctional snf on snf.id=o.sample_Id join cache_taxa_taxon_lists cttl on cttl.id=o.taxa_taxon_list_id where o.created_on>now()-'1 month'::interval