No, the district would *not* be rejected. The errant page number would be reported as "before" page number, that is a number that is before the range determined. The algorithm first finds a group of page numbers that have an occurrence at least half the maximum occurrence of any page number. If such a group contains too large a gap (now set to larger than 4) or has too many gaps (currently more than 20%) it "rejects" the district. It then tries to extend the range at either end with page numbers that are contiguous. After the next update the number of gaps found the in the range will also be reported, that is as well as the current "before/after" figures. Barrie -----Original Message----- From: Philip Powell [mailto:philip@blencathra.org.uk] Sent: 14 January 2004 18:07 To: FREEBMD-DISCUSS-L@rootsweb.com Subject: Re: Missing districts? In message <3F66BB5AC5E4D511B23C00902746C443FAA557@WWMESSD207>, Archer Barrie <Barrie.Archer@services.fujitsu.com> writes >The algorithm that works out the page range has various limits that are >designed to ensure that there is a reasonable degree of confidence in >the page range produced. One of these is that there is not a gap of >greater than 1 page number in the sequence. In Dec 1884 Deaths for >Marylebone,1a there is a gap between page numbers 374 and 377, i.e. two >page numbers missing. > >This limit was particularly aimed at situations at either end of the >range and where there were relatively few page numbers to work with. We >will ease the limit so that it does not exclude the cases with good >populations of pages. Another related problem that I raised elsewhere [I think it was just with Dave] is with regard to possible page number typos. West Derby, in the period I'm most used to transcribing, has a normal page range c800-900. A West Derby entry mistype as 85 would presumable cause the page range algorithm to reject/ignore West Derby even though it is one of the largest districts? I mention this simply as something that people might need to be aware of and not as a criticism. -- Philip Powell Looking north across the Derwent Valley and Northumberland to The Cheviot

In message <3F66BB5AC5E4D511B23C00902746C443FAA563@WWMESSD207>, Archer Barrie <Barrie.Archer@services.fujitsu.com> writes >No, the district would *not* be rejected. The errant page number would be >reported as "before" page number, that is a number that is before the range >determined. > >The algorithm first finds a group of page numbers that have an occurrence at >least half the maximum occurrence of any page number. If such a group >contains too large a gap (now set to larger than 4) or has too many gaps >(currently more than 20%) it "rejects" the district. It then tries to extend >the range at either end with page numbers that are contiguous. > >After the next update the number of gaps found the in the range will also be >reported, that is as well as the current "before/after" figures. Thanks for the clarification. -- Philip Powell Looking north across the Derwent Valley and Northumberland to The Cheviot