Graphical anomaly in Excel 2000

My graphs of linear data appear non-linear just left of
the point of inflection, i.e., appear to curve convexly.
For instance, in a linear plot of (x,y) pairs, just to the
left of the data point where the slope changes from zero
to negative (or positive) the line segment connecting the
data points curves noticeably. This problem appears to be
most noticeable when the data fall on a gridline.
While the values for the dependent variable arise from a
formula I have copied the cells as values only and when
these were graphed, the graph has the same anomaly. Any
help with this vexation would be appreciated.

Howard -

Strange phenomenon. Could you paste a sample of this data (tab
delimited text is fine) so we can try to understand better what you are
describing.

- Jon
-------
Jon Peltier, Microsoft Excel MVP
http://www.geocities.com/jonpeltier/Excel/index.html
_______

Howard wrote:
> My graphs of linear data appear non-linear just left of
> the point of inflection, i.e., appear to curve convexly.
> For instance, in a linear plot of (x,y) pairs, just to the
> left of the data point where the slope changes from zero
> to negative (or positive) the line segment connecting the
> data points curves noticeably. This problem appears to be
> most noticeable when the data fall on a gridline.
> While the values for the dependent variable arise from a
> formula I have copied the cells as values only and when
> these were graphed, the graph has the same anomaly. Any
> help with this vexation would be appreciated.

Howard wrote:
>> My graphs of linear data appear non-linear just left of
>> the point of inflection, i.e., appear to curve convexly.
>> For instance, in a linear plot of (x,y) pairs, just to the
>> left of the data point where the slope changes from zero
>> to negative (or positive) the line segment connecting the
>> data points curves noticeably. This problem appears to be
>> most noticeable when the data fall on a gridline.
>> While the values for the dependent variable arise from a
>> formula I have copied the cells as values only and when
>> these were graphed, the graph has the same anomaly. Any
>> help with this vexation would be appreciated.

I didn't see the original post, so this will appear out of whack...

Howard, might you be using the smoothed line version of the XY scatter
chart? Select the chart, go to Chart | Chart type on the menu. If your
chart uses the smoothed lines, you might get the behavior you describe.
Pick the icon with the more angular lines between data points to get rid of
the curvature.

I might add that I am assuming that you have piecewise linear data. If it
were truly linear, you would not have a change in slope or any inflections.
If I have misunderstood your message and therefore misdiagnosed your issue,
please post back.

Dave
dvt at psu dot edu

[This followup was posted to microsoft.public.excel.charting and a copy
was sent to the cited author.]

Obviously, your question makes sense to Jon, but not to me. Linear data
do not have a point of inflexion. The slope of a linear graph is a
constant. If you do have non-linear data, you might have selected the
'Smooth curve' option. That, under certain circumstances, causes XL to
create a curve that, to the human eye, appears to be somewhat
exaggerated.

--
Regards,

Tushar Mehta
MS MVP Excel 2000-2003
www.tushar-mehta.com
Excel, PowerPoint, and VBA tutorials and add-ins
Custom Productivity Solutions leveraging MS Office


In article <048601c3809e$158f8060$a101280a@phx.gbl>,
hbowles@centurytel.net says...
> My graphs of linear data appear non-linear just left of
> the point of inflection, i.e., appear to curve convexly.
> For instance, in a linear plot of (x,y) pairs, just to the
> left of the data point where the slope changes from zero
> to negative (or positive) the line segment connecting the
> data points curves noticeably. This problem appears to be
> most noticeable when the data fall on a gridline.
> While the values for the dependent variable arise from a
> formula I have copied the cells as values only and when
> these were graphed, the graph has the same anomaly. Any
> help with this vexation would be appreciated.
>