Try plotting two lines in a XY-chart:
Each line should consist of 2 (X,Y) points. The X values of both lines
should be set 0 and 5. The Y values should be calculated from other cells as:
Y=S1*X+K1 and Y=S2*X+K2. Where S1 and S2 are the slopes and K1 and K2 are
the interceptions. Set S1=1, S2=2, K1=0.001 and K2=0.001. Plot the lines in a
XY-Chart. The next step is important: Set the X-Axes scaled from 0 to 0.1 and
the Y-axes scaled from 0 to 0.1.

The BUG:
Now try changing S1 to S1=4. It is clear that something is wrong.The lines
should have different slope but they have the same!!
Go back, set S1=1. Try to chage the X value of 5 to something else, eg.
100.. whoaa.. strange result. There should have been no change in the chart,
but suddently the chart shows a plot, where both slopes are same but not
equal to what S1 and S2 where in the beginning!


Can anyone help me to sort out the problem? I where trying to plot something
with a small zoom on, but the result where WRONG!
Is this a real bug? Does anyone get the same error?

I've left an example on the net:
http://home.tiscali.dk/phreaky/excel....chart.bug.xls


--
Philip

With XL2003 I cannot reproduce the first problem. However, I can the
2nd. Wonder if it is because of a floating point rounding/truncation
error. Don't know what to do with it, though. I wonder if you can
check if the code at xlrotor.com (that should duplicate XL's chart
lines) duplicates the problem.

FWIW, add a trendline with equations and the 2 equations are shown in
the same physical location and cannot be separated. Obviously,
something strange is going on with those extreme values, but I have no
idea what.

--
Regards,

Tushar Mehta
www.tushar-mehta.com
Excel, PowerPoint, and VBA add-ins, tutorials
Custom MS Office productivity solutions

In article ,
says...
Try plotting two lines in a XY-chart:
Each line should consist of 2 (X,Y) points. The X values of both lines
should be set 0 and 5. The Y values should be calculated from other cells as:
Y=S1*X+K1 and Y=S2*X+K2. Where S1 and S2 are the slopes and K1 and K2 are
the interceptions. Set S1=1, S2=2, K1=0.001 and K2=0.001. Plot the lines in a
XY-Chart. The next step is important: Set the X-Axes scaled from 0 to 0.1 and
the Y-axes scaled from 0 to 0.1.

The BUG:
Now try changing S1 to S1=4. It is clear that something is wrong.The lines
should have different slope but they have the same!!
Go back, set S1=1. Try to chage the X value of 5 to something else, eg.
100.. whoaa.. strange result. There should have been no change in the chart,
but suddently the chart shows a plot, where both slopes are same but not
equal to what S1 and S2 where in the beginning!


Can anyone help me to sort out the problem? I where trying to plot something
with a small zoom on, but the result where WRONG!
Is this a real bug? Does anyone get the same error?

I've left an example on the net:
http://home.tiscali.dk/phreaky/excel....chart.bug.xls

I tried it in XL 2000, XP, and 2003 with the following results:

1) The lines are not coincident, but series1 is also not right.
Corresponding to y=0.1, you should see x=(0.1-K1)/S1=0.02475. The
placement varies slightly among versions, but is in the vicinity of 0.035

2) Your results are reproduced in all three versions.

I had previously observed that zooming in on the scale with smoothed
lines could produce very inaccurate curves between points, but had not
known that the problem also plagued straight connecting lines.

Jerry

Philip_plf wrote:

Try plotting two lines in a XY-chart:
Each line should consist of 2 (X,Y) points. The X values of both lines
should be set 0 and 5. The Y values should be calculated from other cells as:
Y=S1*X+K1 and Y=S2*X+K2. Where S1 and S2 are the slopes and K1 and K2 are
the interceptions. Set S1=1, S2=2, K1=0.001 and K2=0.001. Plot the lines in a
XY-Chart. The next step is important: Set the X-Axes scaled from 0 to 0.1 and
the Y-axes scaled from 0 to 0.1.

The BUG:
Now try changing S1 to S1=4. It is clear that something is wrong.The lines
should have different slope but they have the same!!
Go back, set S1=1. Try to chage the X value of 5 to something else, eg.
100.. whoaa.. strange result. There should have been no change in the chart,
but suddently the chart shows a plot, where both slopes are same but not
equal to what S1 and S2 where in the beginning!


Can anyone help me to sort out the problem? I where trying to plot something
with a small zoom on, but the result where WRONG!
Is this a real bug? Does anyone get the same error?

I've left an example on the net:
http://home.tiscali.dk/phreaky/excel....chart.bug.xls

Tushar,

This indeed is very strange. I played around with the sample file, and
observed many quirky things. Enough to drive you nuts.

My guess is that what's behind the strange behavior are the points that
are actually far off the plot because of the manually set axis scales.

When I developed the bezier code I noticed that excel tries to get
clever when plotted points are off scale. Especially when smooth
curves were turned on. I was expecting that "smooth curves" was
involved here, as well. But it's not.

Brian



--
xlrotor
------------------------------------------------------------------------
Posted via http://www.mcse.ms
------------------------------------------------------------------------
View this thread: http://www.mcse.ms/message1411111.html

Hi Brian,

Thanks for your comments. I guess this is yet another 'limit' on how
well XL's charting module handles different user requirements. {g}

--
Regards,

Tushar Mehta
www.tushar-mehta.com
Excel, PowerPoint, and VBA add-ins, tutorials
Custom MS Office productivity solutions

In article ,
says...

Tushar,

This indeed is very strange. I played around with the sample file, and
observed many quirky things. Enough to drive you nuts.

My guess is that what's behind the strange behavior are the points that
are actually far off the plot because of the manually set axis scales.

When I developed the bezier code I noticed that excel tries to get
clever when plotted points are off scale. Especially when smooth
curves were turned on. I was expecting that "smooth curves" was
involved here, as well. But it's not.

Brian



--
xlrotor
------------------------------------------------------------------------
Posted via http://www.mcse.ms
------------------------------------------------------------------------
View this thread: http://www.mcse.ms/message1411111.html

Hi Tushar

Take also a look at the trendlines. The boxes which display the equations
can be moved when the lines look okey. The boxes can for some reason not be
moved when the lines are plotted wrong. They stay fixed in the upper right
corner no matter what you do.

Yep, some crappy code they put behind that drawing routine.

Do you think MS will correct this problem in a future release?

-----
Philip

"Tushar Mehta" wrote:

Hi Brian,

Thanks for your comments. I guess this is yet another 'limit' on how
well XL's charting module handles different user requirements. {g}

--
Regards,

Tushar Mehta
www.tushar-mehta.com
Excel, PowerPoint, and VBA add-ins, tutorials
Custom MS Office productivity solutions

In article ,
says...

Tushar,

This indeed is very strange. I played around with the sample file, and
observed many quirky things. Enough to drive you nuts.

My guess is that what's behind the strange behavior are the points that
are actually far off the plot because of the manually set axis scales.

When I developed the bezier code I noticed that excel tries to get
clever when plotted points are off scale. Especially when smooth
curves were turned on. I was expecting that "smooth curves" was
involved here, as well. But it's not.

Brian



--
xlrotor
------------------------------------------------------------------------
Posted via http://www.mcse.ms
------------------------------------------------------------------------
View this thread: http://www.mcse.ms/message1411111.html

Hi Philip,

Yes, I noted that problem with the trendline boxes in my first post.

As to whether MS will fix the problem, my guess is "doubtful." After
all, it hasn't fixed the problem introduced in XL2002 where one cannot
format (delete) just one segment of a line chart. By comparison, this
is so esoteric that I doubt it would even be considered. Then, of
course, MS might just prove me wrong. {g}

--
Regards,

Tushar Mehta
www.tushar-mehta.com
Excel, PowerPoint, and VBA add-ins, tutorials
Custom MS Office productivity solutions

In article ,
says...
Hi Tushar

Take also a look at the trendlines. The boxes which display the equations
can be moved when the lines look okey. The boxes can for some reason not be
moved when the lines are plotted wrong. They stay fixed in the upper right
corner no matter what you do.

Yep, some crappy code they put behind that drawing routine.

Do you think MS will correct this problem in a future release?

-----
Philip

"Tushar Mehta" wrote:

Hi Brian,

Thanks for your comments. I guess this is yet another 'limit' on how
well XL's charting module handles different user requirements. {g}

--
Regards,

Tushar Mehta
www.tushar-mehta.com
Excel, PowerPoint, and VBA add-ins, tutorials
Custom MS Office productivity solutions

In article ,
says...

Tushar,

This indeed is very strange. I played around with the sample file, and
observed many quirky things. Enough to drive you nuts.

My guess is that what's behind the strange behavior are the points that
are actually far off the plot because of the manually set axis scales.

When I developed the bezier code I noticed that excel tries to get
clever when plotted points are off scale. Especially when smooth
curves were turned on. I was expecting that "smooth curves" was
involved here, as well. But it's not.

Brian



--
xlrotor
------------------------------------------------------------------------
Posted via http://www.mcse.ms
------------------------------------------------------------------------
View this thread: http://www.mcse.ms/message1411111.html