DEMP processing of user map files

description of implementation (preliminary)

This proposed agreement is new. Its terms are subject to revision by the Lawrence Berkeley National Laboratory (LBNL) legal department. If you wish to accept this offer, you will need to sign a statement releasing LBNL from any liability resulting from inaccuracies in the density-equalized maps, or resulting from our delays in producing them. The wording of that statement has not yet been finalized.

LBNL's uncompensated efforts fall in the same category as proposal development, and may not take precedence over currently funded projects. The benefit to LBNL is our enhanced potential for soliciting new funding, through increased public awareness of the DEMP technique,

the proposed agreement

If you will provide a map file and population data in the format specified below, we will send back the corresponding density-equalized map(s). We will provide data files and software for transforming arbitrary points from the original map to the density-equalized map, and vice versa.

We do not guarantee that your map can be successfully density equalized, especially if the population density is highly non- uniform. Minor topological errors are to be expected, for polygons having zero or very small population.

If you plan to use the maps to analyze cases of disease or other discrete events, we do NOT require your individual case records. It is your responsibility to obtain the case data and to develop the statistical techniques for analyzing the density-equalized maps. Please read carefully "Appropriate Use of Density Equalized Maps."

In return for providing this service we require:

  1. your permission for us to publish images of the original map and the density equalized maps (excluding your case data) on the Web. We will credit you publicly, using the information provided in file 1 below. The published images will be in a format that will not permit reconstruction of your original map file.
  2. your permission for us to use these results in soliciting future funding for our own work.
  3. that you credit LBNL in all your published and unpublished work derived from these maps, citing the LBNL URL that we will provide.
If you require more maps or larger maps than specified below, or if you do not want your maps to be publicly available, we are prepared to discuss special funding arrangements. We are also able to assist with development of the necessary input maps and population data.

technical specifications

For computational efficiency, unnecessary geographic detail must be eliminated from the original map. For best results, the average number of points per polygon should be about 5-10. The map should have no more than about 500 polygons. Internal lakes, if present, should be included as polygons. There must be no outlying islands. If a given area, for example a county, has two polygons, the two records must have unique sets of geocodes, and population estimates must be provided for each.

The following four files should returned by e-mail to dwmerrill@lbl.gov, either as quoted text within a single ASCII message, or as quoted ASCII (not binary) enclosures.

file 1: documentation

Please provide the following information:

  1. your name:
  2. full address:
  3. telephone:
  4. fax:
  5. e-mail address:
  6. the URL of your Web home page, if you have one:
  7. the URL of a Web page describing your project, if you have one:
  8. the URL of a Web page describing this file submission, if you have one:
  9. scope of the geographic study area, for example "Alameda and Contra Costa Counties, California":
  10. description of polygons, for example "1980 Census tracts":
  11. npoly = the number of polygons; i.e. the number of lines in file 2 and file 3:
  12. npts = the total number of points in all polygons; i.e. the number of lines in file 4:
  13. x coordinate, for example "degrees longitude (west longitude is negative)":
  14. y coordinate, for example "degrees latitude (south latitude is negative)":
  15. brief description of each geocode field, for example "FIPS county code" or "not used". Indicate who defined each geocode, for example "U.S. Bureau of the Census, 1980".
    geo1:
    geo2:
    geo3:
    geo4:
  16. brief description of each population field, for example "1980 population of non-Hispanic white females age 45-64" or "not used". For each field indicate the source of the data.
    pop1:
    pop2:
    pop3:
    pop4:
  17. A paragraph describing in more detail the source of the map files and population data. Include relevant URL's if available.
  18. A paragraph describing the project in which you propose to use the DEMP map(s) provided by LBNL. Include relevant URL's if available.

file 2: population variables

Exact column locations and field widths are immaterial. Fields on the same line should be separated by one or more spaces, or by tabs (control-I). 
format:
geo1	geo2	geo3	geo4	pts	pop1	pop2	pop3	pop4

example:
6	1	1251	0	4	1247	1.52	15	1.06E+02
6	1	1252	0	7	612	0.	9	2.12E+03
6	2	1412	0	5	0	4.21	12	0.00E+00
etc
File 2 contains npoly lines, one for each polygon. Each line contains four integer geocodes (geo1,geo2,geo3,geo4); the number of points (pts) in the polygon; and up to four non-negative population fields (pop1,pop2,pop3,pop4). The order of polygons in file 2 must be the same as in files 3 and 4.

geo1,geo2,geo3,geo4:
Each line in file 2 (including internal lakes if any are present) must have a unique set of geocodes (geo1,geo2,geo3,geo4). If you use fewer than four geocodes, fill the unused fields with zero.

pts:
The number of points in the polygon (greater than or equal to 3). Do not include repeated endpoints. The sum of the "pts" column in file 2 must be equal to npts, which is the number of lines in file 4.

pop1,pop2,pop3,pop4:
May not be negative. If you use fewer than four population fields, fill the unused fields with zero.

file 3: polygon descriptions

Exact column locations and field widths are immaterial. Fields on the same line should be separated by one or more spaces, or by tabs (control-I). 

format:
geo1	geo2	geo3	geo4	desc		

example:
6	1	1251	0	"Bloomington county"
6	1	1252	0	"Tracyville tract 1252"
6	2	1412	0	"New Paltz township"	
etc
File 3 contains npoly lines, one for each polygon. Each line contains four integer geocodes (geo1,geo2,geo3,geo4), and a character field "desc". The order of polygons in file 3 must be the same as in files 2 and 4.

geo1,geo2,geo3,geo4:
as in file 2.

desc:
The name of the geographic area, surrounded by double quotes. Brief comments may be included (within the same quotes).

file 4: polygon boundaries

The exact column locations and field widths are immaterial. Fields on the same line should be separated by one or more spaces, or by tabs (control-I). 

format:
x	 y

example:
-35.1534 52.3058
-35.5476 52.3022
-35.5432 52.7190
-35.1481 52.7155
-35.7043 48.1565
-35.0104 48.1449
etc
File 4 contains npts lines, one for each point in each polygon. In each polygon, points should be listed in counter-clockwise order. Do NOT repeat the starting point of each polygon. The choice of x and y coordinates is arbitrary, but for geographic data the preferred choice is degrees of longitude and latitude, respectively. Values of west longitude and south latitude should be negative. The order of polygons in file 4 must be the same as in files 2 and 3.

File 4 must be topologically correct, which implies: (a) points shared by adjacent polygons have exactly the same x and y coordinates (b) no two points in any polygon have the same coordinates (c) no polygon has zero area (d) no segments in the map intersect each other (e) the area enclosed by the external boundary is exactly equal to the sum of all the polygon areas.

back to
DEMP software (RLInt)
Density Equalizing Map Projections (DEMP)

usermaps.html 1/14/96
dwmerrill@lbl.gov

http://merrill.olm.net/dwmerrill.html