y-errorbars are no problem. wks.col*.type=3 works. But this only gives
y-errorbars.
Furthermore if I predefine a worksheet with a column specified as x-errorbar,
it is changed to a y-errorbar after import of data.
X-errorbars can of course be defined by hand via the error bar options. But
how
with scripting ? I didn't find a possibility in the manual.

Although one cannot set an error bar column as X error in script, one should be
able to customize a worksheet template (std format: X, Y, Err bar),
setting the error bar column as X Error Bar, and make a call to the template
(errbar.otw for an example) in script and execute the import, i.e.:

win -t data errbar;

getfile *.dat;
open -w %A;

File import routines should not change the error bar type.

thanks for the answer. This is a possible and sensible solution if one does only one graph . For my problem - I read in a set of columns, calculate some new, and plot various of these columns in 5 different graphs. This is done by using the same worksheet and changing the worksheet column types before plotting the next graph it would be rather tedious. I think I would have to open 5 different worksheet templates to make 5 different worksheets with the same data but other column assignments.... This is a way to blow up the projects awfully Nevertheless I think there must be a solution with the set commands: I can set the errorbar color with set %E_deltax -c 2 after plotting, where deltax is the column and 2 is red. I can set the cap width with set %E_deltax -z 20 and the linewidth with -l. If I check the attributes in a script window via get %E_deltax -k;type -b ($(V1))); I can see that the value changes from 3 to 19 if I change the error bar type from Y to X in the error bar menu (a simple wks.col**.type=19 instead of 3 does not work) So I think the changing between Y-bar type and x-bar type must be possible by some simple set command or something like that (I suppose this is simply a set bit in a variable(handleor pointer or whatever) which describes the error bar options. This is not described in the manual

Due to the nature of a powerful, and programmable software as Origin,
there are a few times when certain operations are missed during the
design of the programming language.

Whenever an advanced user runs into such an oversight, we immediately
take action for the next version.

You can count on having a script accessible way of assigning a column
to be Xerr for 5.0.

Keeping with current syntax, the command will most likely be:
wks.col##.type=7;

Thank you for your input.