[Kst] In offset mode, [0] location

Matthew D Truch matt at truch.net
Mon Feb 6 16:45:36 CET 2006

>    I am repeatedly having people ask me if we can align the [0] in offset mode 
> in plots to be exactly at the Y axis, and therefore have the label be exactly 
> what was input, as opposed to putting it inside somewhat.  Why exactly do we 
> move it inward?  And can we change this?

I have thought about this case before.  Note: I did not write any of the
code, nor come up with any of the algorithm, so I might be completely

I believe the [0] is placed 'slightly inward' for the same reason that
the major tick spacing can only be set coarse, normal, fine, etc.  Just
like kst picks the 'right' tick spacing so that the labels are 'nice'
values, kst picks the 'roundest' tick (which are already 'nice') to be
the [0].  This is almost never at the y axis (although it may be), and
may even be quite far to the right.  

Every other plotting program that I have used that has a similar 'base
and offset' mode does what George suggests, that is force the [0] to be in
line with the left y-axis.  The base will no longer neccesarily be a
'nice' value nor neccesarily be the 'roundest' value.  

Benefits of each method:

Current kst 'base and offset mode':
- The base is 'nice' and 'round'.  
- Is what kst has always done.
- When scrolling live data and not reading from beginning, you still
have a sense of the plot axes moving as the data scrolls.

Force leftmost [0]:
- Slightly more intuitive.
- What (most?) other plotting programs do.

Personally, I can see a time and a place for both.  Maybe it should be a
plot option: "Force leftmost [0]".  

"Things should be made as simple as possible, but not any simpler. -- Einstein."
Matthew Truch
Department of Physics
Brown University
matt at truch.net
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
Url : http://mail.kde.org/pipermail/kst/attachments/20060206/3aefb84e/attachment.pgp 

More information about the Kst mailing list