G.    Assumptions

• The Care Giver Center information is not exhaustive and may be incomplete. The same for Drug Stores. This is for time limitation
• The Vacant land information is not real
• The lots cost are not accurate (information previous)

H.    Strategies or Considerations

• 8 or 9 hours of Internet in searching process & EXCEL format preparation. I obtained 112 Care Giver Centers matched  in ArcMap™
• Similar effort & process for 121 Drug Stores

I.        Tools & Information

• ArcGIS 8.2
• ArcMap
• ArcCatalog
• Spatial Analyst Extension. RLIS (Regional Land Information System)
  • FIRE_STA
  • RAILROADS
  • TAXLOTS
  • TRACT
  • VACANT
  • WET_FILL
  • ZONING
• Yellow pages (Addresses & General Information)
• Geocoded services

A.    Commendation on the process of working NEW SITES

1.     I made a special folder in the PC for the project.

2.     I matched 271 addresses of more than 600 I Found in Yellow Pages. The rest were addresses of other parts in Oregon

3.     I TAG every Care Giver Center in accordance at the first letter of its name.            
     

4.     I tag some of the Drug Stores with the Logo of its social reason.     

5.     I generate the raster of straight distance to Rail Roads         

I utilized this 15 step gradient for the zones in this stage. Later I used 10 steps instead of 15

6.     Similar process for the rest of raster distance options [Care Giver,Drug Stores, Fire Station, and Hospitals). I saved these versions in my directory DISK. You can observe different radio shadows because there is transparency given at every raster          

7.     I got the General Zone where every Care Giver Center is actually settled. I believe that the general zone is necessary to be known in order to give a better idea for when I present the addresses of the new Care Giver Centers.

7.1.   I used a new MXD file with ZOONING & CAREGIVER layers

7.2.   I added a new field to Care Giver TABLE (ZONE_CLASS)

7.3.   Step by step I was SELECTING BY LOCATION, AGR Agriculture, COM Commercial, FOR Forest, IND Industrial, MFR Multi-family residential, PUB Public/semi-public, RUR Rural, SFR Single family residential, VAC Undeveloped. While I was modifying mi new field in only those rows SELECTED by location. I succeed in classified 150 Care Giver Center. With the next result 
               

So it means that our 2 new Care Giver Center have to be in SFR Single family residential Zones

8.     I want to consider the population density in my SPATIAL EQUATION. The higher density, better the place. So in TRACT; I divided the population 2000 by 43,560 (an acre in Sq Ft) / area. It means density.         

9.     The final Raster Classification in 10 steps and the Qualification for the formula:        

9.1.   These means that I sum 35, 100, 85, 50, 75, 62.5=407.5. I divided 35*100/407.5=8.59        

9.2.   The Care Giver Center [CGC], Rail Road and Population Density are: The highest value, the better.

9.3.   Drug Store, Hospitals and Fire Station are: The lowest value, the better for us  

10. The must important for the new centers, for me, is the distance to Rail Roads. It has to be at least 5,000 ft away because the noise

11. I reclassified the 6 Raster in order to have the better value and the same scale, the must desirable item         
 

12. I applied the following formulas to COMBINE the rasters:

12.1.                    [DiCareCRecla] * .0859 + [DiRaRoRecla] * .2454 + [DensityRecla] * .2086 + [DiDrugSRecla] * .1227 + [DiHosRecla] * .1840 + [DiFireSRecla] * .1534        

Where:

The result was the RAIL ROADS with a buffer.           
                              
I did not understand the reason. So I tried

12.2.                    [DiCareCRecla] * .0859 - [DiRaRoRecla] * .2454 + [DensityRecla] * .2086 + [DiDrugSRecla] * .1227 + [DiHosRecla] * .1840 + [DiFireSRecla] * .1534        

Gave the same

12.3.                    I took away Rail Road 

[DiCareCRecla] * .0859 + [DensityRecla] * .2086 + [DiDrugSRecla] * .1227 + [DiHosRecla] * .1840 + [DiFireSRecla] * .1534  

Gave this result. It was very logical       
                 

12.4.                    I used again the Rail Road but divided by 10    

[DiCareCRecla] * .0859 - [DiRaRoRecla] * .02454 + [DensityRecla] * .2086 + [DiDrugSRecla] * .1227 + [DiHosRecla] * .1840 + [DiFireSRecla] * .1534

Gave the same Rail Road         

So I decided to utilize 12.3 and, in a posterior stage, consider the Rail Road Distance

12.5.                    I did the calculation in 3 stages: The Distance between Care Centers and the Population Density in step one. So the darkest are the better areas      

12.6.                    The Distance between Drug Stores, Hospitals, Fire Station. So the darkest area, are the better areas    

12.7.                    I utilized the Raster Calculator with these last two rasters. The Darkest area, the most suitable areas      

12.8.                    I tried to rest the Rail Road to both rasters. One more time, the result was illogical. So I will take the Rail Roads into consideration later 

12.9.                    I added Taxlots Only the lots type SFR y VAC

13. I reclassified the Raster for be able to be converted to Polygon SHP           

14. Using the Polygon file. I Select By Location from those lots type VAC that lied in my polygons type 9. I found 1647 lots. I added a field at TAXLOT table to give a mark to those 1,647 lots

15. Here is my consideration to RAIL ROAD  

From those 1,647; I select from TAXLOTS those lots in a distance of 5,000 ft or less. Resulting 1,030. I switched the selection and I got 617 LOTS  

16. I exported to ACCESS that table. I made the report of those lots between 10,000 and 15,000 sq ft in area. The report is the appendices 01. I am reporting only 11 lots from $32,000 to $212,000

B.     Commendation on the process of working QUALIFICATION

In order to obtain the qualification, I felt was necessary to have information from every Care Giver Center distance to:

• Hospitals
• Fire Station
• Drug Stores
• Rail Station
• Witch population density it is lied on     

I was trying to do:

• UNION of lyers,
• Scripts to inherit the RASTER distances values to the TABLE of Care Giver Center.

I Set up a ingenious way of obtaining the values:

1.     I added at the Care Giver Center Table 5 new fields

2.     By   of Care Giver Center

2.1.   I found the max distance between CGC and hospitals. It was 70,000 ft. So I actualized all the rows to 70,000 feet.          

2.2.   I tried again my select, but now with 60,000 ft. And I actualized all the selected rows with 60,000.

2.3.   The same for 50,000 ft. …….. until 1,000 ft. So I obtained a distance scale of every CGC to hospitals.

3.     The same for Firestat, DrugS, Rail For different values:       

4.     In the case of Tract. In Properties of the layer I settled a condition in

So, all the polygons were available to my SELECT BY LOCATION. I ran the Select by Location and I actualize the entire field to cero. I ran my  in Properties of TRACT with  and I ran the Select by Location and I actualize the entire field to 2.9. ….. until I finished in the follow table:     

1.     As we can see in the following table, the data is quite different ranges. Considering too that some data is distance and other is qty/area.    

2.     I converted from ft to Mi. and I gave a Specific Weight at every item. The bigger number, the mayor importance. The sum of all the WEIGHTS is = 100. 

3.     The Sum, Maxim, Minim, Average of every item was quit different. It was necessary to classify the data. So I add the entire sum=2,407. In the tract case, I divided 1,153/2,407= 0.48               

I obtained the 12.53 in tract from divided 6/0.48          

4.     Now, the data is classified. So I am giving importance at the variations of the data. So now I can give qualification at every occurrence. In the TRACT case, I considered negative its contribution :            
In the case of RAIL ROAD I punished in great scale when the distance from CGC to Rail Road was less or equal to 1,000 ft. The second row (0.04) > 3,157 was obtained: from 4*34.48/0.0437=3,156.64

5.     The final step was to sum the data to obtain a final qualification. I give classification, considering the range to give a MAX of 10 and a MIN of 1.         

The entire table is an Addendum at this document

• New Sites searching process was very interesting. The Rail Road raster interpretation, is no total clear for me
• I could not inherit the values of the features in order to give a qualification at every CGC. However I succeed in give the result using alternatives ways.