Dear Jarangote,



While I can't help you with the specific error you are getting, I do want to
add a comment to your posting regarding the use of the K-S
(Kolmogorov-Smirnoff) Goodness-of-Fit criteria for your model to value an
electricity contract.



The K-S statistic is one of the many "Goodness of Fit Criteria" that can
help you decide which distributions fits your data the best. However, we
never use the K-S GOF criteria because it is only concerned with the maximum
vertical distance between the cumulative distribution function of the fitted
distribution and the cumulative distribution of the data. In other words,
the value of the K-S criteria is only determined by the one largest
discrepancy and takes no account of the lack of fit across the rest of the
distribution (i.e. the tails).



The insensitivity of the K-S statistic to lack of fit at the extremes of the
distributions is corrected for in the Anderson-Darling statistic. You
mentioned that you are using it for valuation of an electricity contract, in
which case not just the body of the distribution is important, but I assume
the tails are very important too. The A-D statistic is generally a more
useful measure of fit than the K-S statistic, especially where it is
important to place equal emphasis on fitting a distribution at the tails as
well as the main body. The A-D criteria is therefore, especially in your
situation, a preferred criteria above the K-S criteria. If you want to read
more, see M0255 and M0029 of ModelAssist for Crystal Ball.



Last but not least, from your email I can't determine what data you are
fitting a distribution to. However, in case you are fitting a distribution
to historical electricity prices, you have to be very careful (i.e.
typically fitting a distribution to historical electricity prices is not
correct) since electricity prices are time-series. In other words, when you
are forecasting electricity prices it is important not to do this with a
straightforward distribution-fit, since such approach will likely greatly
underestimate your price-risk. Instead, you need to use a time-series
method.



Hope this helps!

Huybert



From: cbug2@yahoogroups.com [mailto:cbug2@yahoogroups.com] On Behalf Of
jarangote
Sent: Thursday, July 02, 2009 8:17 AM
To: cbug2@yahoogroups.com
Subject: [CBUG] Unable to complete operation due to: Unable to cast object
of type 'System.Doub








Hi,

I have been using Crystall Ball 7 for over a year now which we use to come
up with a valuation of an electricity contract. As new information comes in,
it will be fitted based on the most appropriate distribution table using SK
criteria. I am now trying to input new information but I am getting this
error

"Unable to complete operation due to: Unable to cast object of type
'System.Double' to type 'Excel.Range'."

Can someone help me on this?

Thanks.

J





[Non-text portions of this message have been removed]