extrapolation of lissing data

  • Thread starter Thread starter Guest
  • Start date Start date
G

Guest

I have data sets with data missing that can be filled in with linear extrapolation between the two knowns, though how can this be done
FORECAST only predicts ahead and the missing data isn't always in the same place. Using an IF(ISNUMBER) statement, how can missing data be filled in

1, 2.3, 4, 5, x, x, 5.6, 8, x, 9.5 etc

Thank

Roma
 
I have data sets with data missing that can be filled in with linear extrapolation between the two knowns, though how can this be done?
FORECAST only predicts ahead and the missing data isn't always in the same place. Using an IF(ISNUMBER) statement, how can missing data be filled in?

1, 2.3, 4, 5, x, x, 5.6, 8, x, 9.5 etc.

Thanks

Roman
Click to expand...

Perhaps you could respond to my suggestion of two hours ago instead of posting
a new message?


--ron
 
I have data sets with data missing that can be filled in with linear extrapolation between the two knowns, though how can this be done?
FORECAST only predicts ahead and the missing data isn't always in the same place. Using an IF(ISNUMBER) statement, how can missing data be filled in?

1, 2.3, 4, 5, x, x, 5.6, 8, x, 9.5 etc.

Thanks

Roman
Click to expand...

First of all, FORECAST (or TREND) does NOT only predict ahead. It can equally
well predict values "in the middle", so to speak. So the function can be used
to predict your values 5,6 and 9.

Given this data set, there is another issue with linear extrapolation about
which YOU will need to make a decision.

For example, if your entire known data is used as the "knowns", then the values
for unknowns 5 and 6 will be different than if you assume, for the sake of this
extrapolation, that the only "knowns" are items 4 & 7.

By this I mean that if you look at your values 5 & 5.6 in isolation, then the
extrapolated values in position 5 & 6 will be 5.2 & 5.4.

But if one is drawing a straight line through the entire data set, then the
extrapolated values in positions 5 & 6 will fall along that straight line, and
compute to values of 5.057142857 & 5.921848739. Note that the value in
position 6 is actually greater than the value in position 7.



--ron
 
Back
Top