Figure 6.2. Screenshot of the File menu

This menu allows:

Figure 6.3. Screenshot of the View menu

This menu concerns the appearance of the maps. It provides menu items to zoom in, zoom out and to choose the different panels that can be displayed as different parts of the window:

Depending on the current loaded dataset, the "Display" submenu may also include an additional checkbox item as shown on the Figure 6.4. This checkbox allows the user to display or hide the main cities over the map. By default, if the dataset provides such a layer, it is checked.

Figure 6.4. Display submenu options: cities layer

On this screenshot, the loaded dataset embeds the main cities. The "Display" menu allows to hide or display this additional layer.

When the Display-Cities menu item is enabled, cities are displayed over the maps as black squares, as shown on Figure 6.5. Note that for ergonomy reasons, to avoid overlapping between cities labels, the names of the cities are not displayed over the map. Nevertheless, a tooltip appears when the mouse comes over a square.

Figure 6.5. Displayed cities

Cities are represented as black squares. The name of the city appears when the mouse moves over a square.

Figure 6.6. Screenshot of the Tools menu

The Popup Freeze menu item has been available since HyperAtlas v2. This functionnalitiy is usefull to compare several maps or charts: clicking on this menu item, a popup window is opened, displaying a frozen image of the current visisted tab.

Two options allow to manage the behavious of the mouse cursor:

Other tools available on this menu:

This version 2 of ESPON HyperAtlas allows to define a new study area. On clicking this menu item, the user is invited to enter a name for his/her new study area and to select the top-levels units (as a rule, countries) that will compose this study area.

Figure 6.7 shows the example of a user who wants to define the benelux study area. He/she selects Belgium, Luxembourg and Nederlands units then clicks the "Submit" button. Figure 6.8 shows the information message that is displayed when the creation is successfull. The benelux parameter is now available from the Study Area combo box of the parameters panel. Figure 6.9 shows that interactive maps have been consequently updated on selecting this new study area.

Figure 6.7. Study area creation window

Provides the list of countries and a text field to enter a name for this new study area.

Figure 6.8. Study area creation success

Infomation message.

Figure 6.9. Map of the new study area

Selected new study area.

Figure 6.10. Screenshot of the Session menu

This menu allows to save the parameters of the current analysis to an ESPON HyperAtlas XML file on your disk.

In the case when you already saved such a file, this menu allows to load your previous session parameters.

	A session parameters file is specific to a dataset. An error occurres if you try to load a session parameters file that was built while using another dataset.

Figure 6.11. Screenshot of the Help menu

This menu provides the following items:

Taking account the European level as an example, this section focuses on the importance of considering the multiscalar typologies of regions in political decisions.

When the policymakers want to build political scenari or when they want to evaluate propositions of structural funds, they need to get a synthetic view on the situations of regions which depend on the various territorial contexts.

The question of perequation (transfer from “advanced” to “lagging” region) is very sensible and it is important to propose a complete view of the scales where those perequation processes can take place, according to the principle of subsidiary.

As an example, we analyse how the picture of “lagging” regions is modified when the previous criterion of Objective 1 (less than 75% of the mean value of GDP) is applied simultaneously at three scales: European, national and local.

Simultaneously, it is possible to propose a typology of “advanced regions” based on the symmetric criteria of more than 133% of the mean value of GDP at those three scales.

According to this methodology, it is possible to demonstrate that very few regions are “lagging at all scales” and “advanced at all scales”. Many are in more complex situations, like certain regions of Switzerland or Norway which are “advanced” at European scale, but they are “lagging” at their national or local scales.

Reversely, the metropolitan regions of candidate countries are very often “lagging at European scale” but “advanced at national and local scales”.

The setting of the study area should be the first step of any analysis. Setting the basis of the study can be done by answering the following questions:

Figure 7.1. Study area fields

• Study Area shows the spatial extension that will be mapped.

• Elementary zoning shows the set of elementary units that will be studied.

Figure 7.2. Combination of study area and elementary zoning

These two maps were extracted from the "Area and Zoning" tab of the application with following settings:

	Study Area	Elementary Zoning
Map on the left	European Union 15	NUTS 0
Map on the right	New member states 12	NUTS 3

ESPON HyperAtlas only works with size variables (that is to say that only variables that may be aggregated at upper level by sum), and proposes a multiscalar cartography of the ratio for two size variables in order to set the studied ratio. The user can combine every couple of these types of variables in the “Indicators” box, by choosing each of them in the associated select boxes as shown on Figure 7.3.

Figure 7.3. Indicators box

This box provides three select boxes to choose indicators. The user selected the Density item in the Ratio select box:

• Numerator is set to Population in 2003

• Denominator is set to Area in km²

• Ratio: depending on the chosen dataset (the hyp file), selecting a ratio may implie the auto-selection of the numerator and denominator fields values.

Three maps are respectively linked to these choices, under three different tabs (see Figure 7.4). The maps for the numerator and for the denominator (size variables) are represented with proportional circles. The ratio map is shown with colored units, according to the ratio value. The number of classes and their associated colors (the pallet tool) can be can be set in the "Option" tab of the ratio map.

Figure 7.4. Numerator, denominator and ratio tabs

The three associated tabs to the chosen indicators are represented here for the study area EU 15 (15 countries in Europe) with the NUTS 0 value (countries) for the elemetary zoning:

• the image on the left shows the Numerator map within its associated tab, here, the population in 2003

• the image in the center shows the Denominator map within its associated tab, here, the area in km²

• the image on the right shows the Ratio map within its associated tab, here, the density.

As described in various contexts paragraph, the user has to define the three territorial contexts which respectively set three different levels of spatial observation: global, medium and local. Figure 7.5 shows the select boxes to set these parameters.

	The names of the references have been updated since the previous versions of ESPON HyperAtlas: Global has been renamed General; Medium has been renamed Territorial; Local has been renamed Spatial.

Figure 7.5. Contexts box

The "Contexts" box allows to set three references for their associated deviations: general, territorial and spatial contexts.

The general context may be the whole chosen study area. In such a case, the associated map will be the same as the associated map to the ratio itself. So, the user may choose another general context or a reference value. For instance, in the example of the EU, even if the study area is the 29 potential countries, it may be of interest to observe the spatial differentiations according to another global reference, for instance the global value associated to EU15. For this level, the user may also exogenously enter a value. By default, this value has first been set to the value of the global area.

The territorial context, on the other hand, has to be a geographical zoning that is an aggregation of the “elementary zoning” that was previously chosen.

The spatial context shows which proximity relation will be the basis of the neighborhood’s definition for each elementary unit. That is usually “contiguity”, but it may also be a relationship based on distances since they have been introduced in the hyp file (units that are less than X kilometers far from), or time-distances. Then, each elementary unit value will be compared to the value of its neighborhood.

A set of three maps are linked to these choices (Figure 7.6). The values of the deviations are transformed into global indexes 100. Thus, values may be interpreted in terms of percentage to the reference value. The maps are drawn with double progression frame centered on 100, in order to highlight the regions that are under the reference value (100), and the ones that have upper values.

Figure 7.6. Deviations maps tabs

These screenshots show the three deviations maps tabs for chosen contexts: general deviation on the left, territorial deviation in the center and spatial deviation on the right.

One synthesis map was already available in the previous version of ESPON HyperAtlas, based on three levels and one treshold, it is described in Ternary synthesis map. A new synthesis map has been added to the application since the version 2.0: see Dual synthesis map.

7.5.1. Ternary synthesis map

The three relative positions about contexts are summarized in one synthetic map. The elementary units are classified in eight classes according to their three relative positions.

In order to reduce the whole combinatory of possible cases, from the "Options" tab close to the synthesis map (Figure 7.7), the user must specify which point of view he wants to focus on: the first "Criterion" parameter shows whether the point of view is to underline the regions whose ratio is greater than, or to underline the contextual values, by selecting less than. This choice depends on the studied indicator (see An example of multiscalar typologies of regions section). For instance high values of unemployment rates point out different types of regions than high values of an indicator of resources. According to which regions have to be differentiated (lagging ones or wining ones), the user must chosse the point of view of the synthesis. Then, the user can choose the threshold percentage.

Figure 7.7. Synthesis map options

The map on Figure 7.8 illustrates the eight different configurations of relative position, according to the three previously chosen contexts and parameters. The legend tab gives for each class the descriptions of the contextual positions. The last class (in white) gather the regions that are not concerned by the chosen comparative criterion whatever the contexts are.

Figure 7.8. Synthesis map tab

This screenshot shows the synthesis map tab for the contexts that were chosen in the previous example shown on Figure 7.6.

When “Histogram” is enabled (see Section 6.3 section), the user may represent the three contextual deviations of a selected (clicked) region as an histogram as shown on Figure 7.9.

Figure 7.9. A deviations synthesis histogram for a regiion

This screenshot shows the synthesis histogram of the clicked region named OUEST (West of France). The general deviation of this region is relative to the UE29 general context. The territorial deviation is relative to the NUTS 0 hierarchical context, the spatial deviation considers the contiguity, e.g. the neighborhood of this region.

7.5.2. Dual synthesis map

The dual synthesis map is a new cartographic tool that has been introduced in the version 2.0 of ESPON HyperAtlas. It aims at showing via a chromatic legend the status of territorial units on taking into account two chosen deviations. This section describes the synthetis opportunity that is offered to analysts thanks to this tool.

The legend of the dual synthesis map shown on Figure 7.10 is composed of four main quarters. The values on both axis range range from 0 to 200% and they represent the percentages of a deviation of a territorial unit relatively to a context of reference. The user is invited to select in an options tab the contexts of deviations to be considered for both axis (among the general, the territorial or the spatial context).

Let's consider the following example: the general deviation has been chosen for the horizontal axis and the spatial deviation for the vertical axis. The four main colors of the legend represent the following cases:

• yellow: the global deviation (X axis) is lower than 100% (the average pivot value) and the spatial deviation (Y axis) is upper 100%

• red: both deviations are upper 100%

• blue: both deviations are lower than 100%

• green: the global deviation (X axis) is upper than 100% and the spatial deviation is lower than 100%

Note that the more far from the value 100 the current deviation is, the more intense the color is. Hence a white square in the middle of the legend: this range of values show the territorial units whose both deviations are around the average, 100.

Figure 7.10. Legend of the dual synthesis map

Graduations and quarters of the dual deviation synthesis map legend.

Let's consider now a concrete example on how the dual deviation map can help analysts: the following screenshots decompose as four steps the synthesis about the situation in 2010 according to the European and National averages of unemployement:

• Figure 7.11 shows in red the territorial units whose unemployement rate is above average both at European and National levels:

  Figure 7.11. Dual synthesis map: red units

  
  
  

• Figure 7.12 shows in blue the territorial units whose unemployement rate is under average both at European and National levels:

  Figure 7.12. Dual synthesis map: blue units

  
  
  

• Figure 7.13 shows in yellow the territorial units whose unemployement rate is above average at European level and under at National level:

  Figure 7.13. Dual synthesis map: yellow units

  
  
  

• Figure 7.14 shows the final typology on the complete synthesis map:

  Figure 7.14. Dual synthesis map: final typology

  
  
  

First of all, let's review the available maps tabs and their main focuses:

At any moment, the position of the mouse cursor on the map provides information about the elementary unit that it points. The content of the table depends on the current map, Figure 8.1 shows the case of the synthesis map where are displayed:

Figure 8.1. Details box for the synthesis map

This screenshot shows the "Details" box on the left bottom corner of the application. The user's mouse is over the Guyane. Associated computed values to this unit are displayed in the box.

Except for the synthesis map, a left click anywhere on the map changes the function of the cursor to “Pan”, as long as this option is on (see Section 6.3).

On the synthesis map, the "hand" mouse pointer shows that the histogram function is on. A right click on a region opens its histogram synthesis view (see synthesis as an histogram).

Each map is associated to a set of three tabs that provide tools to control and to understand the cartography. The choices are valid for the current map. Figure 8.2 displays each the "Options" tabs for an indicator map (that shows proportional circles) while Figure 8.3 displays the available options on a deviation map (palett of colors). The user may also set the thresholds for each class. The "Legend" tab displays the bounds of the classes (left), the number of items for each of them (right), and the associated color. The "Explanation" tab displays some general notes about the goal of the current map.

Figure 8.2. Options for proportional circles

The "Options" tab of the numerator or denominator maps aims at setting the representation of the indicator values by selecting a color, the size and transparency of proportional circles.

Figure 8.3. Options for deviation maps

The "Options" tab of the deviation maps aims at setting the representation of the deviation by selecting:

• the palett of colors, that can be reversed

• the number of classes, between two and ten classes

• the progression:

  • arithmetic for classes with an equal amplitude, better choice when the distribution is symmetric around 100.

  • geometric for classes with an increasing amplitude

• the thresholds, that can be edited for each class

It is possible to zoom in/out a map, either on clicking the "View - Zoom" menu items, by using the cursor on the left side of the map or by moving the mouse scroller over the map.

	Please note: the available zoom levels depend on the selected elementary zoning parameter: reduced zoom factor for high levels, maximized at lowest level; any zoom factor update or pan move are applied to every map; the scale of the map is consequently updated.

Figure 8.4. Spatial zoom slider

This screenshot shows the scale, the pan buttons and the zoom slider.

The user can save his current whole collection of maps with the associated rough data and deviations by selecting the "File - Build Report" menu item from the menu bar.

By selecting this menu item, the user is invited to select a directory on his/her disk where the report will be generated as a set of HTML page (index.html and eight PNG image files (one image per map: map0.png, map1.png, to map7.png).

For example, if the user selected his /home/toto/my_hyperatlas_report/ directory as target directory, he/she may open the saved report from a web browser by selecting the /home/toto/my_hyperatlas_report/index.html file.

The generated report may be divided in the three-fold:

	In expert mode, the generated report also includes expert tabs as images: the three equi-repartition maps the tab showing the Lorenz curve and the main statistical indexes the boxplots chart the spatial autocorrelation chart

Figure 8.5. Screenshot of a generated report

This screenshot shows an extract of the generated report index.html file that has been opened by in web browser. This image shows the last map (synthesis) and the start of the table that includes all results.

The map of large deviation provided by ESPON HyperAtlas is a general measure of disparities for a given variable Z which is the ratio between two stocks X and Y. This estimation of general disparities can be further analysed using various econometric indexes that have been added in ESPON HyperAtlas v2 expert mode:

	A complete description of the Lorenz curve and of the main statistical indexes is directly available in a dedicated "Explanation" panel of the statistical box, close to the curve panel, as shown on Figure 9.2.

Figure 9.2. Lorenz curve, statistical indexes and explanations

This tab shows the Lorenz curve, a table of main statistical indexes, and an "Explanations" titled panel providing some information for each feature.

The equi-repartition maps indicate which process of redistribution should be realized in absolute terms in order to achieve convergence, at European, national or local levels.

The equirepartition map is a bi-color discs map showing an absolute deviation. It examines how much amount of the numerator should be moved in order to reach equi-repartition, for each territorial unit, taking into account as a reference the selected deviation context value.

Thus, three equi-repartition maps are available in expert mode for respectively the large, medium and local deviations tabs. As an example, Figure 9.3 shows the equi-repartition map (also called "Redistribution" map) for the large context.

Figure 9.3. Equi-repartition map

Bicolor discs map.

For each unit in the chosen medium context (NUTS 0 for example in Figure 9.4), this chart shows the dispersion of the medium deviation for the territorial units at sub-levels (NUTS 2 level in Figure 9.4).

A boxplot typically provides the following information:

Figure 9.4. Boxplots chart

Available in expert mode.

The spatial autocorrelation chart is only available when the expert mode is enabled.

For each territorial unit of the study area, this chart crosses the values of the spatial deviation on absissa axis with the values of the territorial deviation on ordinates axis.

This chart is very interesting for expert users as it reveals spatial dependancy, e.g. spatial organization of a phenomena.

More empirically, the chart can also be used to examine the situation of outliers and exceptional units out of the cloud of points.

The compute of this chart is based on a Moran's coefficient of spatial autocorrelation variant. The regression line is drawn in red on the chart, its equation, computed by the least squares method, is displayed on the left corner of the frame, as shown on Figure 9.5.

Each unit is drawn as a blue square, its name is displayed in a tooltip when the mouse comes over the square.

Figure 9.5. Spatial autocorrelation chart

Available in expert mode.

This section describes the stocks (statistical data) file that ESPON HyperAdmin expects as input.

Please note the following requirements for the input data file: the input data file must be a spreadsheet xls file (editable by Microsoft Excel and Open Office) named "*data*.xls": the filename must include the "data" characters sequence and the .xls extension is required. the values of stocks must be provided for the lowest level of territorial units. This list is available in the example data template that depends on the selected structure/geometry model at previous step. For your information, the list of units names is given in the appendix Annex: templates units names. all values for all units must be filled;

Following sections describe the expected format (sheets, columns and possible values) for the version 2 of this data.xls file.

12.1.2. Data

Table 12.2 provides an example for this mandatory sheet in the data v2 input xls file.

Table 12.2. V2 sample Data sheet

UT_ID	pop2000	pop2002	area2000	gdp2000	gdp2002
AT111	1	15	2	7	10
AT112	3	16	4	8	11
AT113	5	17	6	9	12

This sheet must provide at least three columns: UT_ID then at least two indicators identifiers (in HyperAtlas, there must be at least one numerator stock and one denominator stock). The Table 12.2 shows five indicators identifiers: pop2000, pop2002, area2000, gdp2000 and gdp2002. These identifiers must be described in the StockInfo sheet (see Section 12.1.8).

The UT_ID column must provide the list of territorial units at the lowest rank (example, at NUTS 3 level) of the dataset. The units are referenced by their identifiers that must match the given values in the associated structure.xls input file.

Then, each other cell provides a value for the given indicator column at the given unit row. For example in Table 12.2, 17 is the value for pop2002 indicator in AT113 territorial unit.

	Each cell must be valuated. Missing values are not accepted here.

12.1.4. Label

Table 12.4 provides an example for this mandatory sheet in the data v2 input xls file.

Table 12.4. V2 sample Label sheet

LABEL_ID	LANG_CODE	NAME	DESC
1	EN	Total population	Total population in thousands
1	FR	Population totale	Population totale en milliers
2	EN	Area	Total area
2	FR	Superficie	Superficie totale
3	EN	GDP	Gross domestic product
3	FR	PIB	Produit intérieur brut
4	EN	GDP/Inhabitant	Gross domestic product per inhabitant
4	FR	PIB/Hab	PIB par habitant
5	EN	Density	Density of population
5	FR	Densité	Densité de population

This sheet aims at providing the internationalized names and descriptions for the indicators and predefined ratios. The LABEL_ID and LANG_CODE provides indexes for this table: for a given label identifier there may be several available translations. Thus, the LABEL_ID = 1 is available in english (LANG_CODE = EN) and french (LANG_CODE = FR) languages. In the StockInfo sheet, each indicator reference a label identifier. As several indicators may be similarly named and described (when an indicator is valuated for several dates), these labels have been exported here.

	The language identifier code must be a valid ISO Language Code. These codes are the lower-case, two-letter codes as defined by ISO-639. Nevertheless, the parser supports upper-cases. You can find a full list of these codes at a number of sites, such as: http://www.ics.uci.edu/pub/ietf/http/related/iso639.txt (2011-03-16).

Note that values in the LABEL_ID column may be referenced from the StockInfo sheet (see Section 12.1.8) and from the RatioStock sheet (see Section 12.1.7).

12.1.7. RatioStock

Table 12.7 provides an example for this optional sheet in the data v2 input xls file.

Table 12.7. V2 sample RatioStock sheet

RATIO_ID	LABEL_ID	NUM_ID	DEN_ID	VALIDITY_START	VALIDITY_END
1	4	gdp2000	pop2000	2000	2000
2	4	gdp2002	pop2002	2002	2002
3	5	pop2000	area2000	2000	2000
4	5	pop2002	area2000	2002	2002

This sheet aims at defining relevant ratios for the HyperAtlas "ratio" combo box parameter. Table 12.7 shows the example of two such predefined ratios, each of them for two different dates:

• the GDP/Inhabitant:

  • in 2000 (second line)

  • in 2002 (third line)

• The density of population:

  • in 2000 (fourth line)

  • in 2002 (fifth line)

Each value in the RATIO_ID column must be unique. Doublons will overwrite the previous found value.

Note that the LABEL_ID references the sames labels for the given pairs of numerator/denominator at different dates (4 for lines 2 and 3, 5 for lines 4 and 5). These labels identifiers must be set in theLabel sheet (see Section 12.1.4).

The values in the NUM_ID column and the values in the DEN_ID column must match the identifiers of indicators that are defined in the StockInfo sheet (see Section 12.1.8).

The values in the VALIDITY_START column will only be considered if the value of the TIME_ENABLED column in the About sheet is TRUE (see Section 12.1.1). Then, one relevant ratio can be chosen in HyperAtlas for different dates. Identically for the values in the VALIDITY_END column. Though VALIDITY_START and VALIDITY_END columns are designed to handle time intervals, setting the same value in both columns makes the ratio associated to a timestamp.

	The expected format for both valididy start/end date fields is currently a year in the yyyy pattern. An input like 2010-12-31 is possible but this version of the application will only take into account the year, that is to say 2010 for this example.