Am 01.11.2020 um 21:49 schrieb Allin Cottrell:
On Sun, 1 Nov 2020, Allin Cottrell wrote:
Ah, I now remember what this was about. We shouldn't do it
user setting a special option (not available yet), but here's the deal.
Suppose I have an array of strings of length equal to the length of the
dataset (most likely, with repetition) and I want to impose these on a
"blank slate" series -- perhaps all NAs.
Although the doc currently says the target series must have all positive
integers. I guess this was implicitly meant as "...or be missing" ?
Then I could do
stringify(x, S, <some option>)
and I'd get a series with m distinct values, where m is the number of
distinct strings in S, with numeric codes set to match. Basically it's a
third way to create a string-valued series, besides importation from a
suitable file and stringification of a series that already holds
suitable numeric codes.
Perhaps this could be achieved without an extra option - again, the doc
says all positive integers for normal use; so perhaps if x is a
zero-valued series this could be taken to mean this alternative usage.
(Assuming that NA still has a separate and sensible meaning which should