[ Login ]

Instructions  

The similarity metric implemented in these pages is version 1.0 of the tripartite similarity metric originally described by R. E. Tulloss (1997. Assessment of Similarity Indices for Undesirable Properties and Proposal of a New Index Based on Cost Functions in M. E. Palm and I. H. Chapela, eds. Mycology in Sustainable Development: 122-143). Its design and the motivation for such a metric are given in the cited paper. In the original paper, the author did not discuss input and output (I/O) of a similarity metric in detail. The paper is restricted to comparisons with other methods, development of a set of principles of similarity computation, and development and demonstrations of the tripartite similarity metric. An implementation such as the present one must include I/O functions. The instructions deal largely with I/O. Input: There are four methods of data entry to the similarity calculator -- two requiring manual data entry and two supporting the upload of previously prepared input files (e.g., the comma separated format (.CSV file) that can be output by Microsoft Excel or by a wordprocessor dumping a table into a plain text format). We provide instructions for each of the modes of use. If you are unfamiliar with the tool, you might want to look at the example of Manual Matrix Entry first. Output: The similarity calculator compares a number of lists pairwise; and for each pair of lists, the calculator produces an index of similarity. These indices are presented in an output matrix that have all the columns and rows labeled with list names. In the cell in which row A and column B intersect (call this cell "A,B"), the measure of the similarity of list A and list B will appear. Since the data in cell A,B is the same as the data in cell B,A, we only fill one half of the output data matrix. The diagonal from upper left to lower right would be filled with indices all having the value "1" (a set is identical to itself). Hence, the diagonal of the output matrix is not populated. We suggest that this page and all those subordinate to it, be cited collectively as Tulloss, R. E. and D. C. Tulloss. current_year. Tripartite Similarity Calculator -- Instructions. www.amanitabear.com/similarity/instructions/. (your_access_date). Contact us: ret@njcc.com

2- Manual Matrix Entry Instructions for use of the manual entry mode for data in a presence-absence matrix. 4- Upload Matrix Instructions for use of the file upload mode for supplying input data in a presence-absence matrix. This is the highest level of automation offered by the Similarity Calculator.

Dealing with Start Up Difficulties Are you having problems with your first use of the simlarity metric? Look here.

© 2003, Amanita-Bear Consulting.
- Last modified: 05/04/2005