Latest Contributions by DanMcCoy

Great!  That's getting us very close. Next step for us is tweaking the relative scoring weight of the House# to bump those scores back down a bit, but still have the numbers closer to 916 score higher than the others.  We're doing some trial and error with the .lot.xml file and trying to use the document in http://www.arcgis.com/home/item.html?id=aeb00de638f3492a93308a4a03183c7d .  If you have any hints for scoring specific to the house number in this method and mapping schema,  we'd appreciate it. Thanks again! Dan ... View more

Thanks Brad.  Would this be in addition to .... or instead of .... the "mapping_schema ref" change suggested previously? ... View more

If you would like to see the geocoding in 10.x improved to include the house number matching like in 9.x, you can vote for the following idea: http://ideas.arcgis.com/ideaView?id=087E000000059oLIAQ There's also some discussion about this issue in this post: Re: Locator won't match close street number in single address locator Dan ... View more

Sure, reducing the min match/candidate scores will show more low-scoring matches, but the issue is how do we increase the score of a candidate that has a house number that is closer to the input address than some other random house number on the same street. For exmaple, if I enter "916 I ST", get back all of the addresses on I ST with the same score... 101 I ST, score 55 1609 I ST, score 55 2322 I ST, score 55 915 I ST, score 55 910 I ST, score 55 401 I ST, score 55 + many, many more We should be able to adjust the scoring rules to promote 910 & 915 to have higher scores than the rest.  Obviously, we don't want these scores to be 100, because they don't match.  However, 910 & 915 are closer to 916 than the others, so the scoring should reflect that. I hope that a 10.x locator style can be customized to change the way that the house number matches to behave more like it did with 9.x style locators. Thanks, Dan ... View more

Peter Hanmore: It's especially annoying that it worked as desired in 9.3.1 and now it's 'broken' in 10. I just wanted to confirm that 9.x style locators using "one address" behave as Peter mentions.  It will match to a house number that is close, but it does drop the score. Dan ... View more

Brad Niemand: there is a way to customize the locator style template file to build a locator without requiring the house number to be correct. What you would do is the following. 1. Open the USAddress.lot.xml file with a XML editor or text editor like Notpad++. 2. Search for "<mapping_schema ref="SingleAddressSearch" />" 3. Replace with "<mapping_schema ref="SingleAddressPolygonCentroid" />". 4. Build the locator again. We tried changing the mapping_schema ref and rebuilding the locator, but did not see any change.  Are there any addtional steps needed to change the scoring weight of the house number component? Thanks, Dan ... View more

We're also getting the  "could not initialize service" error using Direct Connects.  We have an SOE, but it's not running on all of the services that throw the error, and the SOE debugging/logging is disabled. Any other ideas? Thanks, Dan ... View more

Roy Jackson: When you pass in an address without a unit designator to the get candidates method, addresses having a unit are ranked the same as an exact match without a unit.  This effectively means that the geocoding operation does not perform correctly for any location having secondary addresses (has units) AND a primary address (no units), such as an apartment building with a office at the primary address.  Has anyone resolved this issue or know if it's been resolved at 10.3? Thanks, Dan ... View more

Russell Provost: However, if an address is being matched without an apartment number (123 Main Street), but the reference data contains unit information for this address (123 Main Street, Apt. 1 or 123 Main Street, Apt.2) the matching algorithm will return a score of 100 for each instance in the reference data with the same primary address. I cannot solve this behavior in my locator either. Ideally the scoring would be slightly penalized for the matches with unit information. Any suggestions? Has anyone resolved this issue or know if it's been resolved at 10.3? Thanks, Dan ... View more

Please vote for this idea: ArcGIS Idea - Candidates for point address locators ... View more

Peter, I agree with you.... It would make sense to me that you should be able to get a partial match to an address with a similar street number (13 -> 15), but I don't know if it's possible with ArcGIS locators. Were you ever able to find any way of tweaking the scoring or match rules? Dan ... View more

Allister Farn wrote: ESRI have confirmed that they've been able to reproduce this with the latest version of the API. Will wait to hear back from official support channel. Allister, Did ESRI provide you with a BUG#? Thanks, Dan ... View more

We have the same message.  Did you ever find out how to clear it up? Thanks, Dan ... View more