�??US Dual Range Addresses�?� style issue,
I wanted to apply the �??US Dual Range Addresses�?� style but it appears that the searched addresses don�??t match the exact location of the target building.
For example, in the screenshot below, I wanted to search the �??5 peace street, Ramallah�?�. The result doesn�??t match the precise physical location of the building. In other words, how to have the location of the found building just precisely in front of the building. The schema is also shown in the screenshot.
[ATTACH=CONFIG]32846[/ATTACH]
I still need to use the �??Street�?� layer as REFERENCE DATA but I�??m looking for having the precise location of the searched building.
What other options I might still have?
Thank you
Best
Jamal
Jamal,
With a locator based on street ranges, it will never plot exactly on buildings, unless the buildings are right on the street and your ranges are setup perfect.
The closest to perfect you can get is using a composite address locator.
If you use your building layer first and street second. It should first look at your building layer, if an address is found you will get the exact location, if not you will at least get an approximate location by using your street address locator.
Hope this helps!
Tim
Jamal,
The "Dual Ranges" style uses interpolation as an approximation of the position along the street that an address should be. The only style that would be more precise would be the "Single House" style which would require an address point for each house. Is there a reason that the exact location is needed for the geocode? If so, the "Single House" style is your only option and will require a lot of work to add every house to the dataset.
Brad
Thank you very much Brad for the input.
Sure, end users prefer to see geocoded address in the precise place.
[ATTACH=CONFIG]32897[/ATTACH]
Do you mean that the Google database of addresses, for example, is collected at the level of each building as the found addresses match the precise location of the target?
[ATTACH=CONFIG]32898[/ATTACH]
It boils down to what your use is and what your budget is, and how they balance each other: As has been described, anytime you interpolate an address based on a linear feature, you accuracy is okay, but your precision suffers. That's just the name of the beast. Lets say you have a street that is ranged from 100-198 and 101-199 on the even and odd sides respectively. That locator will interpolate house number 150 right in the middle of the block on the even side; the reality might be that its really on the odd side and at the 199 end of the block.
If block level accuracy is 'good enough' then your are good to go. However, if you need better precision than that (like the actual location of house #150) you'll need different data with which to match. Accuracy might cost you $X, while Precision might cost you $X * 5. However, what is the result of an accurate location versus a precise location? That's something only you can answer.
Thank you Joe for the useful answer.
In my case, I wanted to explore if there is a more innovative approach to use the �??Dual Range�?� style but at the same time to find the EXACT location of the geocoded address.
I've known people to address range their streets to the actual physical addresses. Personally I think that's a complete waste of time and energy. I mean, if you have the actual physical addresses, why not create and use points?
The thing is, using a linear feature is always going to be using a linear feature. It's cheap and easy and there's lots of data out there to use. But it has it's limitations and no matter how hard you try, those limitations will prevail.
I have a colleague who made a great comment in a similar discussion: "What if the Census Bureau /Tiger/DIME files started as points instead of lines? Do you think things would be different today?" Jamal I know you are not in the US but way way way back in the day, Tiger files were all we had to work with. We've come a long way in data quality, but a line is still a line.