MapForums

I have a dataset with 2093 records, using multiple address fields and postcodes. I have imported this and matched those from the choices presented my Mappoint 2004. I noticed an unexpected datapoint at about latitude 54.07 north and 2.78 West. On further inspection I found that 73 addresses had matched to this point.
What is special about this point? All the addresses have Postcodes which are spread throughout the country and nowhere near this point. And these addresses are, as far as I can tell, all addresses that were mapped directly and so regarded as unambiguous. I suspect it may be the geographic centre of the United Kingdom. But how could mappoint be happy with a match just to the country when it has a postcode?

Hi nick

If your address input is simply UK this is where it’s plotted – somewhere near Morecambe if I recall. If you right click on the pushpin file name you create in the legend, then properties, then matching I think it will show you that these 73 addresses were matched by country – basically MP isn’t finding your address sufficiently unambiguously so this is what it’s giving you. Unfortunately MP doesn’t default to the postcode so I normally use a little bit of programming to plot by address, then plot by postcode and then highlight those with a big divergence (normally deferring to postcode on those). Quite often the postcode is more accurate anyway because I think addresses along streets are ‘interpolated’ along that street if that’s the right word. Try re-plotting just based on postcodes by suppressing the rest of the address information (in the Import Data Wizard set the address names to ‘other data’. )

Best rgds

__________________
David
MapPoint Europe Gallery at http://www.broomanalysis.plus.com/gallerylist.html