Re: [Gretl-devel] grabbing Alfred

On Tue, 20 Aug 2013, Riccardo (Jack) Lucchetti wrote: > On Mon, 19 Aug 2013, Allin Cottrell wrote: > >> On Mon, 19 Aug 2013, Allin Cottrell wrote: >> >> So here's a suggestion: when we determine that a certain column of a >> CSV file represents a string-valued variable, by default we treat >> all non-blank values as string literals. But we provide a "set" >> variable ("missing_string" or some such) so that the user can >> specify a missing-code for string-valued input. E.g. when reading >> from Alfred's tab-separated files one could say >> >> set missing_string "." >> >> (This would not just be for "join", but for any delimited-text >> read.) > > We already have "set csv_na"; we could use that. That occurred to me too, but I think it would be confusing. The (only) role of "set csv_na" at present is to set the string used to represent NA on _output_ of CSV from gretl. This string may be specific to the anticipated use of gretl's CSV output by a third-party program (e.g. Stata or Ox, which have their own notions of what represents NA). But what we're talking about now is the string to interpret as NA on _input_ of string-valued variables from CSV. This string may be specific to the third-party producer of the CSV file (e.g. Alfred). The string that's appropriate for one of these uses is not necessarily appropriate for the other use; or to put it differently, "set csv_na" (used as per its current role) could have unwanted side-effects if it were also to govern the reading of string-valued variables from CSV subsequently.