From kde-devel  Fri Feb 21 08:48:02 2003
From: Vladimir Dergachev <volodya () mindspring ! com>
Date: Fri, 21 Feb 2003 08:48:02 +0000
To: kde-devel
Subject: kspread slowness with a lot of data
X-MARC-Message: https://marc.info/?l=kde-devel&m=104581741722395


I would appreciate any comments/advice on the following problem:

I have a CSV file with about 12000 lines and 4 columns. I would like to
make a chart using data from one of columns as X and from another as Y.
I would be nice if Y axis was logarithmic.

Here are the things I tried (with Kspread 1.2.1) :

         * kspread and OpenOffice Calc. Both are *very* slow and I was
           able to get a chart with Calc (but not kspread)

         * kspread actually feels better than OpenOffice Calc when
           scrolling (OpenOffice is "uniformly" slow)

         * there is a large delay when saving kspread data in native
           format. (99% cpu consumption)

         * there is an exact same delay (with 99% cpu consumption)
           when doing many other operations:

                   o creating a chart in a separate sheet
                   o right-clicking on a chart in a separate sheet
                   o some highlighting operations

Problems:

      * I do not know how to tell KChart which data to use

      * I do not know how to populate 12000 cells in a column
        with =log(Ex) (with x varying through 12000 numbers)
        - clicking on a dot and dragging takes too long to
        paint 12000 cells.

      * I would like to get rid of the annoying delay. It looks very much
        like something Kspread calls often "just-in-case" and it degrades
        perfomance with a lot of data. Perhaps this is garbage collection
        of some sort or verification that all cells are up to date ?
        I would appreciate a hint on where to look.

Thank you !

                       best

                           Vladimir Dergachev

 
>> Visit http://mail.kde.org/mailman/listinfo/kde-devel#unsub to unsubscribe <<