On Tue, 9 Feb 2010, Michael Stefaniuc wrote: [...]
It seems the data is sliced three ways:
I needed some catchy titles; I know those aren't that great. If somebody has better ideas I'm all ears.
I don't know if this is any better than your titles, but here's a try:
* Most frequent messages by report Description: Messages that are present in the most reports.
* Most frequent messages by line Description: Messages with the highest cummulative line count that are also present in at least 1% of reports. The latter criteria is to eliminate messages that merely have a very high repetition count in a handful of reports.
* Messages with highest repetition count Description: Messages with the highest repetition count in a single report.
Also:
* I think 'log' would be a good name. In any case I'd only use one name throughout, including in the column heading.
* I'd always leave the columns in the same order, no matter which one the data is sorted on. The reader should pretty easily figure it out and will have less trouble comparing the tables.