1950 ORER (boxcar distribution)


Richard Brennan <brennan8@...>
 

At 07:46 PM 12/30/2008, Mike Aufderheide wrote:
<snip>
Based on the Gilbert/Nelson idea of boxcar distribution it should be
possible to come up with a list of cars most likely to appear at a
given time and place. My goal is to parce through the list and come
up with a list of must-have cars. For example there were about
714,000 boxcars total in 1950. I want to have a 200 boxcar fleet on
my layout so I need to know which cars represented at least 3,570
cars or 0.5% of the total. I've done the work by hand and want to
add it to the spreadsheet.

Mike...

I'm no statistician... but applying a linear factor to this bothers me.
Were you planning to adjust for some probability of interchange by road and/or commodity type?

Maybe even using: <<http://en.wikipedia.org/wiki/Karmarkar's_algorithm>http://en.wikipedia.org/wiki/Karmarkar's_algorithm> ???


--------------------
Richard Brennan - San Leandro CA
--------------------