Users Guide to the Most Similar Neighbor Imputation Program

Partially inventoried planning area Planning area fully populated with inventory information using global data and estimates of ground data based on "most similar neighbors." { You may order additional copies of this publication by sending your mailing information in label form through one of the following media. Please specify the publication title and number. Abstract ______________________________________________ program is used to impute attributes measured on some sample units to sample units where they are not measured. In forestry applications, forest stands or vegetation polygons are examples of sample units. Attributes from detailed vegetation inventories are imputed to sample units where that information is not measured. MSN performs a canonical correlation analysis between information measured on all units and the detailed inventory data to guide the selection of measurements to impute. This report presents an introductory discussion of Most Similar Neighbor imputation and shows how to run the program. An example taken from a forest inventory application is presented with notes on other applications and experiences using MSN. Technical details of the way MSN works are included. Information on how to get and install the program and on computer system requirements is appended. The MSN Web address is: David Renner is a freelance Computer Programmer in Moscow, ID. He has contributed to several projects at the USDA Forest Service, Rocky Mountain Research Station's Moscow Laboratory related to FVS and its application. He also works at the University of Idaho and as a private forestry consultant.