|
PureBytes Links
Trading Reference Links
|
Thanks Gary. Got it working. Any ideas how to construct different
categories for values away from a trendline? I'm trying to come up with
a way that (arbitrarily) measures the (in)significance of values from
the trendline.
-----Original Message-----
From: Gary Funck [mailto:gary@xxxxxxxxxxxx]
Sent: Saturday, December 20, 2003 2:41 PM
To: omega-list@xxxxxxxxxx
Subject: RE: Excel help (again)
You'll need to create a separate column with the running total of days,
starting
at 0. You can take the difference between two adjacent dates and add
that
into
your running total, and fill the formula down. Then create a chart,
using
your
offsets as the x-labels. To get the trendline, right click on a data
point,
and
select "Add Trendline".
If you want your dates on the x axis instead of the 0..N offset, it is
going
to get to be complicated.
If your data series is trading days and you aren't interested in
calendar
day timeframes, then just assume each trading day is one day and that
your
data is contiguous. Then you can use your dates on the X-axis as is, and
still calculate a trend line, keeping in mind that the unit of time is
trading days, not calendar days.
> -----Original Message-----
> From: cwest [mailto:cwest@xxxxxxxxxxxx]
> Sent: Saturday, December 20, 2003 1:30 PM
> To: 'OmegaList'
> Subject: Excel help (again)
>
>
> I have a file of values and dates with different numbers of days
between
> dates. In Excel, what's a reasonable way to measure the distance of
each
> value from a line of best fit so that I can categorize outliers. I'd
> like to be able to say that certain outliers are more or less
> (in)significant.
>
> I played with the LINEST function in Excel but I can't get the hang of
> it, and I couldn't see how I could plot a regression line.
>
>
|