Rules for checking map connections from Ted
Back to Meeting Nov 3, 2010, Notes: Wed, Nov 3 2010
Subject: Quality Control Check of SMS connections Date: Tue, 09 Nov 2010 18:36:07 -0500 From: Ted Willard <twillard@aaas.org> To: Brian Sweeney <bsweeney@aaas.org>, Francis Molina <fmolina@aaas.org>, John Weatherley <jweather@ucar.edu>, Mike Wright <mwright@ucar.edu> Hi all- As promised, here are some "things to look for" in trying to rectify inconsistences between Atlas 1 and Atlas 2 as represented in the Strand Map Service. Issues that need to be examined and/or rectified: 1. Benchmark A and Benchmark B appear on Map X and Map Y. They are connected on Map X but not connected on Map Y 2. Benchmark A and Benchmark B appear on Map X and Map Y, A --> B on Map X but B --> A on Map Y 3. Across all maps Instances where "A --> B --> C --> A" or "A --> B --> C --> D --> A" or "A --> B --> C --> D --> E --> A" and so on. 4. A case of A --> B where A is in a higher grade level than B is. A couple notes: Any time a conflict is found, I will need to know on which maps each of the connections is found in order to work through a solution. There will be instances for number 1 above that are ok, and require no correction, but it would be a good to examine as part of the quality control process. A program that finds the recursive loop "A --> B --> C --> D --> A", there would also be a loops for "B --> C --> D --> A --> B" "C --> D --> A --> B --> C" "D --> A --> B --> C --> D". It would be nice to have the search consider this one problem rather than four separate ones. Finally, there are cases where we have decided when doing Atlas 2 and Benchmarks Online to split a benchmark that appeared as whole in Atlas 1, and also where we took what was two Benchmarks in Atlas 1 and combined them to form one benchmark in Atlas 2 and benchmarks online. I am not sure how the strand map service is currently dealing with these issues. Let me know if you have any questions or concerns. -Ted