Hi Gavin,

I agree that having a warning would be very helpful.  I've found a
workaround, though, that's helpful for me and might be helpful for
you, too.  In the import data window, under "General Options", I add a
prefix (or suffix) that signifies which file the data is coming from.
I make the prefix unique, and thus my imported data never gets
overwritten.

Steve

On Thu, Apr 5, 2012 at 1:47 PM, Junkshops <[email protected]> wrote:
> Hi Jeremy and all,
>
> I'm wondering how difficult it would be to add dataset namespaces. Currently
> if a file is imported with a data row/column label that's the same as a data
> label in a previously imported file, the new data row/column will clobber
> the old one from the previous file. In the use case of multiple experimental
> conditions resulting in files with the same format, that means the
> researcher has to go into each file and add some sort of unique id to the
> label of each row of data if he or she wants to display multiple datasets in
> the same Veusz graph. Additionally, if two imported files usually have
> different labels but in rare cases have the same label, one of the data sets
> will clobber the other without the researcher knowing it (no error or
> warning occurs - the old data is silently overwritten).
>
> Hence if the data labels from one file could have namespaces designated by
> the user, perhaps defaulting to the filenames, it could be very useful for
> researchers dealing with data in multiple files. I'm not sure how much
> reworking of the code that might take though - could be substantial. Anyway,
> just a suggestion - but I would at minimum advocate warning a user if his or
> her data is going to be overwritten.
>
> Cheers, Gavin
>
>
>
>
>
>
> _______________________________________________
> Veusz-discuss mailing list
> [email protected]
> https://mail.gna.org/listinfo/veusz-discuss

_______________________________________________
Veusz-discuss mailing list
[email protected]
https://mail.gna.org/listinfo/veusz-discuss

Répondre à