If you don't care about the spike, clip it.
If you care about the spike, but need to see the movement of the closer points, clip it at the top of the chart range you want to display, and put a * and note the actual value.
If you care about the spike, but only as a relative jump, you could average the surrounding 2 or 3 points into each plotted point, to get a smoother graph showing trends rather than actual points.
NOTE: I have never had a stats course, so any of this could be completely against common practice.
Jerry Johnson
>>> [EMAIL PROTECTED] 07/12/04 09:52AM >>>
Exactly,
-r
On Mon, 12 Jul 2004 09:46:28 -0400, S. Isaac Dealey <[EMAIL PROTECTED]> wrote:
> > I have a simple line graph that represents the number of
> > sessions per hour.
> > In my stats, I have one data point that is significantly
> > different
> > from the other points. Does anyone know a nice way to
> > handle data like
> > that? I'm thinking perhaps of just filtering any data
> > point that is
> > 200% higher/lower than the average, but was curious to see
> > if there
> > was a better way.
>
> You're talking about removing a "spike" from the graph?
>
> s. isaac dealey 954.927.5117
>
> new epoch : isn't it time for a change?
>
> add features without fixtures with
> the onTap open source framework
> http://www.sys-con.com/story/?storyid=44477&DE=1
>
>
[Todays Threads] [This Message] [Subscription] [Fast Unsubscribe] [User Settings] [Donations and Support]
