Classes
Many analyses allow you to assign color, size, or other display properties to illustrate your data by generating classes. This involves summarizing data in groups (so-called classes), and assigning the same display styles to each group.
Note: If you have inserted an analysis via the control window data, three color classes are automatically stored. You can change this in the global Options.
Numeric and Alphanumeric Classification
The data type determines the evaluation of the classes. You can see the data type in the selection list of the Column ("A" stands for alphanumeric columns,"1" for numerical columns). If you have selected a numeric column, the classifcation is based on numeric principles. However, if you have selected a column with the data type "Text", the evaluation is carried out lexically (according to dictionary sorting).
Note: If, for example, a column that actually contains numbers is inadvertently not marked as a number column during data import, easymap will treat it like "text" and sort numbers lexically! For example, a class from 1 to 2 would then contain the value 15!
Methods for Creating Classes
Statistics knows numerous methods of class formation. In easymap you can use the most common ones. The following methods are available for creating classes:
Here the classes can be freely defined by the user.
Let easymap determine the limit values of the classes in such a way that about the same number of data records fall in each class.
This classification type is not only possible when using number columns, but also when using text columns as the data basis. For each occurring data value easymap forms a separate class. Of course, this only leads to meaningful representations if the number of occurring data values is manageable. A frequent application of this classification type is the presentation of the A-B-C customer analysis.
Let easymap determine the class limits in such a way that all classes have the same width (i.e. the difference between upper and lower class limits is the same for all classes).
easymap selects the average value as the limit value of the classes. Of course this is only possible with two classes. The interpretation of this class definition is simple: the lower class contains the below-average data, the upper class the above-average data.
Note: In earlier easymap versions (up to 10.0) this classification type was called "mean value".
The option Sum equal leads to classes whose class sum is approximately equal. This means that all values falling into one class sum up to a similar sum as the summed values of the other classes. The potential or data volume contained in the individual classes is therefore the same. As a result, classes with higher values contain only a few data records, while classes with lower values contain many data records (to arrive at the same sum).
This Classification option is perfect for determining "Priority" areas. You can highlight, for example (preferably few) areas where 25% of sales has already been achieved. On the other hand, such areas can be neglected, because 75% of sales can be achieved with only few areas while the remaining 25%, would have to be added to a very large area.
Example: Pareto distribution (80-to-20 rule)
The "80-to-20-rule" states that most often 20% of market coverage is sufficient to cover 80% of the total volume/ Revenue, e.g.: with just 20% of customers, 80% of sales can be achieved, or by operating 20% of market areas, 80% of customers’ needs are met.
You can solve such problems by a clever classification. Select 5 number of classes. Use the Same Sum classification. Select for the first class value a not deep color, for example, white. Use for the other classes significantly different colors.Already in the classification dialog box, you can see the rule (Pareto Distribution) applied on the available data. The first couple of class values should contain at least 80% of all cases (Frequency column). If you display the analysis now, the areas that you need to at least get 80% of sales are shown in the map. The white areas show the remaining 20% of its sales.
For the classification type normal distribution (Gauss), the class boundaries are aligned using the parameters of the Gauss normal distribution (mean value μ and standard deviation σ). The class width is set to the standard deviation. If the number of classes is even (recommended), the limit between the two mean classes is set to the mean value. If the number of classes is odd, the mean class is centered on the mean value. For two classes, the option normal distribution (Gauss) results in the same classes as the option mean.
In the case of an even number of classes generated with the Normal (or Gaussian) distribution option, the following applies for normally distributed values (provided the classes are completed):
- Approximately 68% of all values fall into the two middle classes
- Approximately 95% of all values fall into the four middle classes
- Over 99% of all values fall into the six middle classes
With the classification type normally distributed (Gauss) easymap always sets the lowest class open downwards and the highest class open upwards. Thus, all values are always classified; the outermost classes contain the extreme values or outliers.
Tip: If you wish to create only closed classes, raise the number of classes by two before applying Normal (or Gaussian) distribution and subsequently delete the lowest and highest classes.
Note: With automatically generated classes, it can happen that classes are omitted due to the data situation. If you have put a lot of effort into the properties of the classes (for example, defining the design characteristics, descriptions), we recommend that you set the classification type to user-defined so that nothing can change in the class structure. However, dynamic adaptation to data changes will then no longer take place.
The class list
The class list displays each class as a row. There is always a remainder class that determines the display of unclassified values (these are all values that do not fall into any of the defined classes). Furthermore, there is always at least one correct class.
If an automatic class calculation is active (that is, a classification type other than user defined) and you make manual changes to the classes, the classification type automatically jumps to user defined.
Class List Columns
Display features | The columns displayed in the class list are variable. With differentiated analyses, only one design characteristic (e.g. color) is always displayed in the first column; with simple analyses, all design characteristics concerned (2-3) are displayed in the first 2-3 columns. The design characteristics can be edited by double-clicking in the respective cell. |
class limitations |
Normally there is a From and a Up to column where you can set the interval that covers the class. While the first value still falls into the class, the last value does not. However, if you use the classification type Attribute, only one column Value is displayed. A data value falls into this class if it is equal to this value. You can also define open classes. In this case, the interval is limited to only one side. To do this, enter * in the cell or delete the value. You can also open both boundaries, so that all values fall into this class. |
Frequency | This shows how many data records falls into each of the classes for the current setting. |
Sum | This displays the sum of all the data records that fall into each of the classes for the current setting. The column is only displayed for summarizable data and not when using the classification type Attribute. |
Caption | You can specify an alternative name for the class that can be used in legends. |
Note: You can also enter multi-line texts. You can use one of the following shortcuts to insert a line break: Shift+Return, Ctrl+Return or Alt+Return.
Sorting the Classes
The columns of the class list can be sorted to change the order in legends. The classes are then also processed in a different sequence.
A data value is always assigned to the first matching class!
Analysis area
Normally, easymap considers all data values (i.e. the value range from minimum to maximum) when creating equally distributed and equidistant classes. However, if you do not want to consider extreme values when creating classes, you can do so by specifying an analysis range. The use of an analysis range is also useful for defining numerous equidistant classes that are graded as integers. For example, to define 20 classes in hundreds starting from zero, simply select the analysis range from 0 to 2000, 20 classes, and then select Equidistant for classification.
Edit classes
The various layout attributes of classes, the boundaries of the individual classes or the caption of the classes for the legend can be adjusted individually. To do this, simply click in the field that you want to change and select a different symbol or enter a different value for the class boundary.
Note: To change several objects at the same time, press the Ctrl key and select the corresponding fields in the list.
For editing and saving classes, colors, size/width etc. easymap also offers various functions that simplify class editing.
Load classes | Load the saved settings from another workbook. |
Save classes | You can save your settings for a class (class range limits, colors, class labels, etc.). |
Smooth Class Range Limits | Here the set class limit values can be rounded up or down. Please note that the set classification is then changed to User defined and the frequencies are recalculated. |
Allow Gaps | Use this function to set the class limits in such a way that the last value in a class is not the first value of the next class. In doing so, individual values can be removed from the display. |
Share classes | The command Share class splits the selected class into two can also be used to easily create symmetrical classes around the mean value. To do this, first select the classification "Average", then divide the two classes and obtain four classes, where the limit value between the second and third classes is the mean value. |
Combine classes | Using the Merge Classes command allows you to create a new class, by automatically merging several existing classes. At least two classes must be selected before executing this command. As a result, the classification mode automatically switches to User-Defined, and frequencies are immediately recalculated. |
Remove classes | You can use this function to delete individual classes. Please note that you must first select the class to be deleted and only then select the command. You can also delete the class directly from the keyboard (key Delete). The values of the deleted class are added to the Not classified class. |
Copy Classification | Copy the classification directly to the clipboard. |
Insert Classification | Insert the classification directly from the clipboard. |
Select symbol set... |
If you have many classes that are to be represented by different icons, you can use predefined symbol sets. This saves the manual selection of symbols for each class. To select a particular symbol set, click on the "symbols" above the class. Some symbol sets are optimized for a certain number of classes, so that when there is a more or less number, a message appears asking whether the number of classes should be adjusted or kept. |
Save symbol set... | You can also save symbols that you have assembled yourself. |
Reverse order | Use this command to reverse the sequence of symbols. |
Load Colors... |
Load saved colors. easymap offers not only the possibility to save and reload your own color palettes, but also a whole range of ready-made color palettes to choose from. To apply one of the color palettes for the display in your surface coloring or the Boston Grid analysis, click on Edit Colors in the menu Edit Colors and select a color palette from one of the directories. You can find these color palettes in the subdirectory of easymap. You have the choice between diverging, qualitative and sequential color palettes. Special color palettes for the Boston Grid analysis can be found in the Boston Grid Matrix subdirectory. You can get a quick overview of the color combination in the Explorer - just set the view in the respective directory to thumbnail. |
Save Colors... | Save the set colors as a file for further use. |
Edit Color Gradient... | Here you can set the set color gradient by specifying two or three colors. Please note that the setting with three colors is only possible from three classes. |
Generate Random Colors... | Here you may generate a mixed color palette setting a certain brightness and saturation. |
Reverse Color Gradient | This command reverses the selected gradient. |
Create Asymmetrical Color Gradient... |
With the command Generate asymmetric gradient you can assign the middle color of the gradient of type "from / over / to" to any class. This is useful when the class boundaries are asymmetric (e.g. -100 to below -10 / -10 to below +10 / 10 to below 200 / 200 to below 500 / 500 and more). Note: If you want an asymmetric gradient, proceed as follows: Select Edit gradient and then Gradient with specification of 3 colors . Then select the class to which you want to assign the middle gradient color, and then choose Generate Asymmetric Gradient . |
Set color effect and transparency... | Here you can adjust the transparency of the colors using a slider or by entering a value. |
Edit Size Gradient... | You can use this function to determine the size for the smallest and largest class. |
Reverse Size Gradient | This function allows you to reverse the size progression with a single click. |
In certain situations, easymap displays warnings in the value class list to alert you to possible unintended situations:
- The value for "From" must be smaller than the value for "Up to": No value would fall into this class. You should correct the class boundaries.
- There is more than one class whose lower/upper limit is open: In this case, the first class found would always "win" for very small or very large values.
- The value range of this class overlaps the value range of another class: This can only happen if "Allow Gaps" is activated. The warning indicates that the class intervals of two classes overlap. In this case, the first class identified would always "win".
Save and load classes
In the menu classes you can save your settings of the classes (class boundaries, colors, class names etc.) and load them again for further analyses. This is useful, for example, if you need to create a series of similar, comparable maps with different selections. Such a class file (*.xml) can be reloaded with Load classes....