Forecast using 2 data sets

[Guest]

I am trying to predict future values using 2 data sets. I'm not sure if
Forecast is the right function or if something else is better suited for
this. My situation could best be illustrated using this simplified example.

I have 2006 data for weeks 1 thru 6. I'm trying to predict future 2006
values for weeks 7 thru 13. I alse have historical data for weeks 7 thru 13
from last year (2005). I would like to predict weeks 7 thru 13 this year
using the preceding 6 week time frame, but also want to use the corresponding
time frame from last year to factor in the appropriate seasonality trend.

Can someone point me in the right direction as to how to set this up? Below
is my attempt at drawing out the data that I have available to use.

wk1 wk2.....wk6 wk7 wk8......wk13
2005 9 8 10
2006 10 11 13 ? ? ?


Thanks,

Jeff

 

[Mike Middleton]

Jeff -

It is unlikely that a "simplified example" describes enough of your problem.

The first step should be to plot the historical data to visually identify
patterns.

A second step might be to select an appropriate model. In your situation
this might involve trend (linear or quadratic) and seasonality (using
indicator variables, lagged variables, or classical time series
decomposition index numbers).

After fitting a model to the historical data, finally prepare predictions of
future values.

Two example XLS files are available at http://www.mikemiddleton.com. These
examples do not have step-by-step instructions; more detailed descriptions
are in Ch. 20 Time Series Seasonality of my book Data Analysis Using
Microsoft Excel: Updated for Office XP. But the examples may give you some
guidance.

- Mike