I want to find out the mathematical formula describing the best
fitting curve for a given set of data (Shipping Cost). I have four
tables (four shipping methods) containing Shipping Costs for each
WeightXContinent. The dependent variable is the shipping charge for a
given weight range (rows), with less than 50 weight ranges distributed
among the four tables. Each table contain 5 columns for any given
weight range, each column representing a destination country group
(continent)

Ideally i want to have the least possible different formulas for all
tables by plugging into the formula some kind of
coefficient value related to the shipping method (and/or continent). I
want to use this formula to
calculate approx shipping costs, with minimal losses across most
situations. Our average
package weight is 2500g and we most often ship to Groups III and IV.

I would also like to know which type of regression is most appropriate
for this situation, and a simple why explanation.

You can download the XLS file from http://tinyurl.com/an8ca

Thank you for your time.

---

Clarification of Question by taxq-ga on 15 Feb 2006 13:56 PST

if the above link doesnt work try
http://osbarcellos.com/dl/WeightXContinent_Shipping_Costs.xls