"Stacked scatter plot"?

[sylvan]

Hi. I want to have a stacked area plot to show off how much time out of the
total various tasks take in some simulation software I have. The problem is
that the input data (the x axis) occurs at irregularintervals. E.g. I may
have data for 10 elements, and then the next is for 20 elements, and then one
data point for 21 elements or whatever.

Just producing a stacked area plot looks allright, but even though the
input:time releationship is more or less linear, I get this weirdly shape
curve because Excell just assumes that the input data value pairs should be
evenely spaced on the X axis.

Is there any way of getting it to put the "ticks" on the X axis depending on
the actual value rather than just uniformly distributing it?

 

[Bernard Liengme]

Tell us more details about the two sets of data
Then tell us what the chart should show.

 

[sylvan]

Like I said, I have various "times" that all add up to a total time. E.g.
"For input of size N, task A takes 4ms, task B takes 6ms, task C takes 1 ms"
etc. Each of these task's times are in a separate column, where each row
corresponds to a different N. I'd like a stacked area chart for this to show
off the total time, and the individual sub times.

This works if I just choose "stacked area" chart, the problem is that my "N"
values are not spaced out uniformly (e.g. N could be 1, 2, 4, 10,
25,50,56,63) so the "ticks" on the X-axis shouldn't be uniformly spaced out,
bur work like a scatter plot.

 

[Jon Peltier]

This may work. Chart menu > Chart Options > Axes tab, change Category (X)
Axis to Time Scale. Then double click the axis, and on the Number tab,
choose General or another relevant non-date format.

- Jon

 

[sylvan]

I'm sorry, I'm using Office 2007 and can not find these options. I did find
the option of setting the X axis to "date" which did indeed spread out the
sample points like they should in the graph, unfortunately it still spaced
out the "ticks" uniformly which is a problem for me (Only certain values for
"N" are actually even valid, so if I have values for N=5 and 7, it's not
acceptable to get a tick at 7, even if the data is correctly fitted to the
actual sample points).