Overview

Each occurrence of the MAPS command produces a single map displaying from one to six themes (bars, pies, dot densities, symbols, and shadings for ranges or individual values) that illustrate the distribution of data across the geographic regions displayed on the map. The map boundaries and geographic features, such as highways and city locations, come from a set of tables known as a geoset. The values of a geographic variable in the SPSS data must match values of a key field in a geoset in order to place the thematic elements in the right geographic regions. Note: Region is used throughout this document to refer to any geographic unit. In fact, most themes can be applied to points, such as cities or office locations, and to lines, such as highways, as well as to areas with boundaries, such as countries.

Basic Specification

The basic specification has three required parts: The name of a geoset. The name of the geographic variable whose values correspond to those of a table in the geoset. (See the XY subcommand for an alternative.) A theme subcommand that includes the variable on which descriptive statistics are to be calculated for each region.

Syntax Rules

One and only one of the GVAR, XY, or LOOKUP subcommands is required to specify the SPSS variable to be matched with a table in the geoset. XY and LOOKUP also provide information to create a new layer. The GSET subcommand is required. At least one of the theme subcommands (ROVMAP, SYMBOLMAP, DOTMAP, IVMAP, BARMAP, or PIEMAP) is required. Each of these can be entered once and only once. The GVAR (or XY or LOOKUP), GSET, LAYER, LOOK, SHOWLABEL, TITLE , and GVMISMATCH subcommands can be entered in any order but must precede the theme subcommands.

Operations

Each MAPS command creates a single map. Data are aggregated to the level of the values of the geographic variable. After aggregation, data values are matched by the values of the geographic variable to the values of a layer in a specified geoset. This is known as data binding. By default, the software looks for a layer whose values match the values of the geographic variable specified on the MAPS command. If multiple themes are requested, they are drawn in this order: individual values, range of values, dot density, pie, bar, and symbol.

Limitations

A maximum of ten bars or pie slices can be shown. For bars corresponding to separate variables, the limit is six. A maximum of 99 values is allowed in an individual values map. Each theme can be applied only once to each map. All themes on a map must be bound to the same layer. For example, it is not possible to have a range of values based on countries and graduated symbols based on cities.

GVAR Subcommand

The GVAR subcommand requires the name of an SPSS variable that identifies the geographic regions, such as COUNTRY or COUNTY. The values of this variable must match the values in a table of the geoset. Occasionally, the values of a single variable do not fully identify regions, as in the case of United States counties, which can occur with the same name in more than one state. In this case, a second variable is required to refine the match.

MAPS /GVAR = VAR(county) VAR(state) /GSET = 'United States' /DOTMAP= VAR(sales) SUM=(sum).
Because the same county name can occur within different states, the variable STATE is required to ensure that COUNTY is unique.

XY Subcommand

The XY subcommand is useful when the SPSS data contain the coordinates of points to be shown on a map. By naming these coordinates, you can create a new layer in the geoset that contains the points and displays themes at those points. This subcommand requires three variables, giving in order the x (longitude) and y (latitude) coordinates and a key variable that identifies the points. The data are aggregated on the key variable; if there is more than
one instance of each value of the key variable in the file, the x/y coordinates are taken from the first occurrence of that value in the data. (The assumption is that all occurrences of the same key value, such as the identity of an office at a particular location, will have the same x/y coordinates.)
MAPS /XY = VAR(x) VAR(y) VAR(company) /GSET = 'United States' /SYMBOLMAP= VAR(sales) SUM=(SUM).
Each company in the data file has unique coordinates, designated x and y. (If some companies had more than one location, it would be necessary to have a variable that designated each location so that all locations would be shown.) A new layer named Company (XY) is created in the geoset. The total (sum) of sales to each company will be represented in the size of a symbol at each of the x/y points.

LOOKUP Subcommand

The LOOKUP subcommand extends the capability of the XY subcommand. It allows you to use coordinates from an existing table to create a new layer in your geoset. For example, if you have zip codes in your data but no x/y coordinates to represent zip codes on your map, and your geoset does not contain a zip code layer, you can instruct SPSS Maps to look up the coordinates in a table and create a new layer, just as in XY binding. In this case, you provide the name of the variable that you want to match to geographic coordinates and the name of the file that contains those coordinates. The data are aggregated on that variable and then matched to values in the lookup table (exactly as geographic variables are matched), and the resulting layer is included in the geoset. The lookup file can be any table in the MapInfo format to which data could be bound. The layer constructed by LOOKUP contains points only for points present in the data, not for all points that might be present in the lookup file. Therefore, the LOOKUP subcommand can be useful whenever you want to create a layer containing just the points of interest to youa selection of cities, perhaps, instead of all of the cities in a geoset layer.

MAPS /LOOKUP = VAR(zip) 'C:\\Program Files\spss10\Maps\ZIPCODE.TAB' /GSET = 'United States' /SYMBOLMAP= VAR(sales) SUM=(SUM).
The SPSS data file contains the zip codes in the variable ZIP. The file ZIPCODE.TAB contains zip codes and the x/y coordinates of their centroids. A new layer containing the coordinates of each zip code in the SPSS data file is added to the geoset.
The total (sum) of sales to each zip code will be represented in the size of a symbol at each of the x/y points. If multiple cases have the same zip code, they will be summed to give the total sales per zip code.

GSET Subcommand

The required GSET subcommand names the geoset that supplies the boundaries, points, and other geographic features for the map. The filename refers to a file with a.GST extension that includes references to the various tables that make up the geoset.

LAYER Keyword

By default, the Maps procedure searches all of the registered tables in the geoset to find one whose values match the values of your geographic variable. It is possible for more than one table in the geoset to contain matching values. You might, for example, have a layer of major cities and another layer of capital cities, with a good deal of duplication between them. The optional LAYER keyword on the GSET subcommand allows you to specify a particular layer in the geoset to which you want your geographic variable to be bound. To find the names of all the layers in a geoset, run the Geoset Manager, which is available from the SPSS for Windows software group on the Start menu.

SHOWLABEL Subcommand

The SHOWLABEL subcommand allows you to specify whether labels are displayed on your map for the layer that matches your geographic variable.

AS_IS NO YES

Displays or hides the labels depending on the setting within the geoset. This is the default. Hides the labels. Displays the labels.

TITLE Subcommand

The TITLE subcommand specifies a title for the map. The default title is the name of the geoset. The title is limited to a single line. Enter the title enclosed in quotation marks or apostrophes. Title attributes (font, size, color) can be changed through editing in the Viewer but cannot be set through command syntax.

GVMISMATCH Subcommand

When a data value in your geographic variable does not match a value in the layer to which it is being bound, a mismatch occurs and a warning is written to a mismatch table in the Viewer. GVMISMATCH allows you to specify the maximum number of mismatches that will be reported. The existence of a value in the geoset that is not in the SPSS data does not constitute a mismatch. If, for example, you do not have data for one of the countries shown on your map, that country will simply appear without a theme in the color and pattern established for it in the geoset.

MAPS /GVAR = VAR(city) /GSET = 'United States' LAYER = 'US Cities' /GVMISMATCH MAX = 50 /IVMAP= VAR(SALESREP) SUM=(MODE).
This map identifies each city with the sales representative who appears most often on the records for that city. The GVMISMATCH subcommand allows up to 50 mismatches to be reported in a warning table. Sales to cities not included in the U.S. Cities layer of the geoset will not be shown on the map.

ROVMAP Subcommand

A range of values map divides the values of a variable into a set of ranges and assigns each geographic unit to one of the ranges. On the map, the ranges are represented as gradations between a color representing the lowest range and another color representing the highest range. Data are first aggregated so that each geographic unit is represented by one case, and then ranges are determined and cases are assigned to ranges.

VAR(varname)

The variable whose ranges are shown on the map. $COUNT can be used instead of VAR(varname) to produce ranges based on the count of cases within each geographic unit. This specification is required. The aggregation to be performed on the specified variable before ranges are determined. Not required if the variable is $COUNT. The method used to distribute cases into ranges. Five methods are available: EQSIZE divides cases into ranges of approximately equal size. EQCOUNT puts approximately the same number of cases in each range. NATBREAK uses an algorithm to distribute data evenly among ranges based on the average of each range. Values in each range are close to the average for that range. SD uses the standard deviation. The middle range breaks at the mean of the data values. The ranges above and below the middle are one
(SUM=function) DISTRIBUTION
standard deviation above or below the mean. CUSTOM allows you to specify your own ranges with the XRANGE keyword.

XRANGE=(n,n)

For custom ranges, specify XRANGE once for each range. Ranges may not overlap. Optionally, you can specify a name for each range, as in XRANGE=(13,19) ' Teenagers'. Whether empty ranges should be allowed. The specifications are YES and NO, with NO being the default for all distribution methods except CUSTOM. With custom ranges, this specification is ignored. The number of ranges to create. Ignored if the distribution method is SD or CUSTOM, or if the number and distribution of cases is too small to produce the requested number of ranges. The title for the legend. (DEFAULT) explicitly requests the default, which is the label of the variable whose ranges are shown, or blank if counts are shown. Determines whether the theme is visible when the map is initially drawn. The default is YES. The alternative, NO, is useful on multipletheme maps where you intend to experiment with which themes to show.

ALLOWEMPTY

NUMRANGES=n

LEGENDTITLE

VISIBLE
MAPS /GVAR = VAR(country) /GSET = 'World Countries' LAYER='World' /TITLE = 'Population Increase' /ROVMAP = VAR(pop_incr) SUM=(MEAN) DISTRIBUTION = SD LEGENDTITLE = ''.
This command generates a map showing the various ranges of population increase in the countries of the world. The SPSS data file contains only one record per country, so no real aggregation takes place. MEAN simply yields the one value per value of COUNTRY. The distribution method is SD, so that ranges of population growth will be one standard deviation wide, with the middle range breaking at the mean.

SYMBOLMAP Subcommand

A graduated symbol map places a symbol on or within each region. The size of the symbol is proportional to the value of a summary function calculated on a single variable within each region.
The variable whose values determine the symbol size for each region. $COUNT can be used instead of VAR(varname) to produce symbols based on the count of cases within each geographic unit. This specification is required.

(SUM=function)

The aggregation to be performed on the specified variable to produce the values represented by the symbol sizes. Not required if the variable is $COUNT. The title for the legend. (DEFAULT) explicitly requests the default, which is the label of the variable represented by the symbols, or blank if counts are shown. Determines whether the theme is visible when the map is initially drawn. The default is YES. The alternative, NO, is useful on multipletheme maps where you intend to experiment with which themes to show.
MAPS /GVAR = VAR(country) /GSET = 'World Countries' LAYER='World' /SYMBOLMAP= VAR(gdp_cap) SUM=(MEAN).
This command produces a map in which a symbol within each country is proportional to that countrys gross domestic product. Because the data contain only one record per country, the MEAN summary function simply yields the value for each country.

DOTMAP Subcommand

A dot density map places within each region a number of dots proportional to the value of a summary function calculated on a single variable within each region. Because the dots must be spread across a region, the geographic variable used in a dot density map must correspond to a layer that contains area boundaries. Dots are distributed randomly within each region.
The variable whose values determine the density of dots for each region. $COUNT can be used instead of VAR(varname) to produce dot densities based on the count of cases within each geographic unit. This specification is required. The aggregation to be performed on the specified variable to produce the values represented by the dot density. Not required if the variable is $COUNT. The data value represented by one dot. The specification can be any positive number, including decimal values less than 1. The title for the legend. (DEFAULT) explicitly requests the default, which is the label of the variable represented by the dots, or blank if counts are shown. Determines whether the theme is visible when the map is initially drawn. The default is YES. The alternative, NO, is useful on multipletheme maps where you intend to experiment with which themes to show.

VALUE1DOT=n LEGENDTITLE

MAPS /GVAR = VAR(fromctry) /GSET = 'World Countries' LAYER='World' /TITLE = 'Total Messages Per Country' /DOTMAP= $COUNT.
This command creates a map that uses dot densities within the borders of each country to show the number of e-mail messages received from that country. The data for this example are records of individual e-mail messages. The geographic variable is the country from which each message originated. The $COUNT stand-in variable requests that the messages be counted for each country.

IVMAP Subcommand

An individual values map uses color and/or pattern differences to indicate the value each region has on a single variable.
VAR(varname) (SUM=function)
The variable whose values determine the color and/or pattern for each region. This specification is required. The aggregation to be performed on the specified variable to produce the values represented by the individual colors. Required even if the data contain only one record per region (in which case you can use any of the functions that return the single value, such as MEAN or MODE). From the dialog boxes, only MODE is available. Not required if the variable is $COUNT. The title for the legend. (DEFAULT) explicitly requests the default, which is the label of the variable whose values are shown. Determines whether the theme is visible when the map is initially drawn. The default is YES. The alternative, NO, is useful on multipletheme maps where you intend to experiment with which themes to show.

LEGENDTITLE VISIBLE

MAPS /GVAR = VAR(country) /GSET = 'World Countries' LAYER='World' /IVMAP= VAR(climate) SUM=(MODE).
This command produces a map in which each country is colored to indicate its predominant climate. The legend contains the value labels for CLIMATE. The MODE function produces the most frequently occurring value for each country. Because this data file contains only one record for each country, that value is obtained and shown.

BARMAP Subcommand

A bar chart map can display bars for multiple variables or for categories determined by a BY variable.

VAR(v1) VAR(v2).

Variables for individual bars. You can list up to six scale variables in the form VAR(varname) VAR(varname). The data are aggregated within the values of the geographic variable; each bar represents all of the cases within each region. See VAR(v1) BY VAR(v2) for the alternative. You can also use $COUNT, but that must be the only variable.
V1 is the variable to be summarized within the bars; you can use $COUNT instead of VAR(v1). V1 must be numeric. The values of v2 divide the data into separate bars. V2 can be numeric or string and

VAR(v1) BY VAR(v2)

should have no more than 10 distinct values.
The aggregation to be performed on the specified variable to produce the values represented by the bars. Not required if the variable is $COUNT. The height for the bar that represents the largest value encountered in the data. The default is 0.25 inches (0.64 cm.). When set to YES, each bar is scaled independent of the other bars so that bar heights can be compared between regions but not between bars in a single chart. This is useful for showing variables measured on different scales, such as population and revenue. The default is NO so that all bars on the map use the same scale. The title for the legend. (DEFAULT) explicitly requests the default, which is blank if more than one variable is represented in the bars or if counts are shown and otherwise is the name of the variable whose values determine the heights of the bars. Determines whether the theme is visible when the map is initially drawn. The default is YES. The alternative, NO, is useful on multipletheme maps where you intend to experiment with which themes to show.

HEIGHT INDSCALE

MAPS /GVAR = VAR(state) /GSET = 'United States' /TITLE = 'Sales by Size of Customer' /BARMAP= $COUNT BY VAR(cosize3).
This command produces a map of the United States with a bar chart in each state indicating the number (count) of individual sales made to small, medium, and large customers within each state. The data are records of individual sales.

MAPS /GVAR = VAR(country) /GSET = 'World Countries' LAYER='World' /TITLE = 'World Literacy Rates' /BARMAP= VAR(lit_fema) VAR(lit_male) SUM=(MEAN) INDSCALE=NO.
This command creates a world map and places a bar chart on each country showing the female and male literacy rates. Because the data contain only one record per country, the MEAN summary function yields that value for each country. INDSCALE=NO is the default, included here for illustration. Because the same scale is used for both variables, the bar heights allow you to compare relative female and male literacy rates within each country. If it were YES, then both female and male literacy rates would be relative to that in other countries but independent of each other.

PIEMAP Subcommand

V1 is the variable to be summarized within each pie; you can use $COUNT instead of VAR(v1). V1 must be numeric. The values of v2 divide the pie into slices. V2 can be numeric or string and should have
no more than 10 distinct values. Both variables are required.
The aggregation to be performed on V1 to produce the values represented by the slices in each pie. Not required if the variable is $COUNT. The diameter of each pie. If GRADUATED=ON, this is the diameter of the largest pie. The default is 0.25 inches (0.64 cm.). When GRADUATED=YES, the diameters of pies within the map are scaled according to the total value represented by the whole pie, enabling comparisons between regions. The default is YES. The title for the legend. (DEFAULT) explicitly requests the default, which is the label of the variable that determines the size of the slices (V1 in the description), or blank if counts are shown. Determines whether the theme is visible when the map is initially drawn. The default is YES. The alternative, NO, is useful on multipletheme maps where you intend to experiment with which themes to show.

DIAMETER GRADUATED

MAPS /GVAR = VAR(state) /GSET = 'United States' /TITLE = 'Sales by Customer Type' /PIEMAP= VAR(sale_prd) BY VAR(industry) SUM=(SUM) GRADUATED = YES LEGENDTITLE = ''.
This command produces a map of the United States with a pie chart in each state indicating the sum of product sales by customer type (industry). Because GRADUATED=YES, the pies are scaled so that their diameters are proportional to the total sales for each state relative to that of the other states. The null legend title prevents the variable label for INDUSTRY from being printed there, since the title is used to give that information.

Summary Functions

The following functions are available for any map theme. Some may be inappropriate, such as means and standard deviations in pie charts or individual values charts, and are not available through the graphical user interface, but you are not prevented from using them in the command language. To obtain counts, use $COUNT in place of VAR(varname) as indicated in the sections on theme subcommands.

First Values. The value found in the first case for each category in the data file at the time the summary function was assigned. Last Values. The value found in the last case for each category in the data file that created it. Maximum Values. The largest value within each category. Means. The arithmetic average for each category. Medians. The value below which half of the cases fall in each category. If there is an even number of cases, the median is the average of the two middle cases when they are sorted in ascending or descending order. Minimum Values. The smallest value within each category. Modes. The most frequently occurring value within each category. If multiple modes exist,
the smallest value is used.
Number of Cases Above (N of Cases >). The number of cases having values above the specified

value.

Number of Cases Between (N Between). The number of cases between two specified values. Number of Cases Equal to (N of Cases =). The number of cases equal to the specified value. Number of Cases Greater Than or Equal to (N of Cases >=). The number of cases having values
above or equal to the specified value.
Number of Cases Less Than (N of Cases <). The number of cases below the specified value. Number of Cases Less Than or Equal to (N of Cases <=). The number of cases below or equal to

the specified value.

Standard Deviations (SD). A measure of dispersion around the mean, expressed in the same unit of measurement as the observations, equal to the square root of the variance. In a normal distribution, 68% of cases fall within one standard deviation of the mean and 95% of cases fall within two standard deviations. Sums. The sums of the values within each category. Variances. A measure of how much observations vary from the mean, expressed in squared

units.

Subject Index
Administrador del geodiccionario, 16, 18
enlace de datos, 15 opciones para los mapas temticos, 32
barras propiedades en mapas, 80
funciones de resumen en mapas, 6
capas adicin al resultado de mapa, 89 cambio de nivel de zoom, 90 en conjuntos geogrficos, 13 modificacin de capas del mapa, 90 modificacin en mapas, 70, 87 propiedades en mapas, 87 seleccin en la creacin de mapas, 32 conjuntos geogrficos, 9 Administrador de conjuntos geogrficos, 11 Administrador del geodiccionario, 16 capas, 10 creacin, 12 discordancia de datos, 94 instalacin, 9 modificacin, 13 registro, 15, 16, 18 seleccin en la creacin de mapas, 23, 26 coordenadas x/y en mapas temticos, 29