As you pointed out, This didn't work properly. I'll fix it at once.
(2) creating histograms:
The initial design was made to be able to cast int to float and later I
changed the behavior for some reason.
But now I also think it's better to get the capability back and will
include into the next release.
Let me apologize for causing you inconvenience of 100% CPU problem.
But I've been working for resolving it and found a way to reduce CPU
and memory usage to some extent.
Please be patient for some more days till the next release.
On 2004/11/09, at 0:08, dmwt@... wrote:
> Hi Koji
> thanks for the fixes and sorry for my late answers...
> Indeed now it opens and closes the trees from my files properly. Only
> 100%CPU problem remains which occasionally appeasrs when working with
> on the pyhton command line and browsing the history of commands and
> the cursor. Very strange but lets forget about that. Maybe it is my
> Unfortunatly I don't have more than one computer where I can test
> Now my scripts (currently I write scripts for JAIDA/jython and just
> run them
> under PAIDA to see what they do) get stuck at some other problem:
> (1) when opening a file with these options
> I think it should recreate a file from scratch even if it exists (at
> thats what JAIDA seems to do). PAIDA just opens it and if you try to
> some histo which already exists it throws an exception. I tried to
> this by setting "setOverwrite()" on this tree but it didn't help. also
> stuff with rmdir did not work?!
> (2) creating histograms:
> if you do it like hf.createHistogram1D("bla","bla",10,0,10) it will
> an exception because it does not want integers as boundaries.
> hf.createHistogram1D("bla","bla",10,0.,10.) is fine but it is annoying
> have to do the trivial int to float casting yourself.
> Thats it for the moment. I discovered that PAIDA is the only tool to
> 3D histos and histos with variable bin widths properly; so I would
> like to
> abandon jython and JAIDA at some point (if execution speed is