This is either a slippery slope we should be wary of tumbling
down, or a problem that needs a fuller solution, not just a partial
fix. I constantly interact with spreadsheet users who lay out a
worksheet [?is this the right terminology, I don't use spreadsheets really]
with quite a few lines of text (titles, column headers, blank lines)
before what Stata would regard as the data, strict sense. And some of this
text is likely to be free form text that is not intended as Stata variable
names or variable labels, or not mappable to such. (Commonly the spreadsheet
is managed with complete ignorance of Stata, often understandably so.)
This all makes perfect sense from the point of view of spreadsheet use -- you can put
your "metadata" where you like in readable form -- but is a pain for
reading into Stata. So it can be much more than a matter of what is contained
in the first and second lines.
The uncompromising solution is to remind users of Stata's rules and
put the onus on users to clean up. There is more guidance at
http://www.stata.com/support/faqs/data/newexcel1.html. It is ironic
that I am credited as lead author, as that FAQ is based mostly
on a compilation from other people created because I got questions
from local users that I often couldn't answer myself.
A fuller solution might entail options for indicating what should
be done about blank lines, title lines and header lines.
Blank columns are another detail, although the fix inside Stata
is easy.
Nick
[email protected]
David Kantor
> Thank you for that information and the suggestion. I would put on my
> wishlist for -insheet- to have an option to handle the
> var-labels-in-second-line situation.
>
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
