If you have recorded data on several species of a genus, you can automatically create a summary of these data. This could provide a description of a genus, which applies to all species of this genus included in the data set. You can use also this feature to create equivalent summaries for feature groups. You could, e. g., be interested to see a summary of all items living in a certain habitat.

You must select a source project, from which the items to be summarized are retrieved, and a destination project, to which the newly summarized items will be written. The source and the destination project can be identical.

If you enter the name of a non-existing project as destination project, a new project will be created. The character definition of the new project will be linked to the source object and the item definition and description will be created as an empty table.

To summarize items, you must first create a table defining how items shall be summarized. If the source project name is "X", the name of this table will be "X_Item_Summarize". Not that although you have several options to automatically create this table, you can manually modify the "SummaryName" attribute to change which items will be summarized later on.

Summarize based on offers the following options:

• complete ItemName: Use this option if you have multiple items with identical item names. This happens if you have created separate items to record data from voucher specimen or different literature references. This is recommended, because only in this way you can later correct errors introduced by misapplied names, or remove some literature information, which you found to be unreliable. The summarized items will be equivalent to taxon (usually species) descriptions.
• first word in ItemName ("genus"): Use this option if you want to create generic descriptions from species. In the absence of a taxonomic subsystem, DeltaAccess has no idea of nomenclature and does not really know about a genus, but the first word of the item name usually is a good choice. The first word is defined as all characters up to the first blank, period, comma, ";", "(", or "[".
• the following character...: Use this option to summarize items according to the character you selected. If the character is categorical, the new summarized items will use the character state names as item names, if it is a text character, the text is used directly. Not that if a categorical character is not exclusive, a single item may be added to several "character-state-items", i.e. the sum of items may be larger than your original items.

Press the button Create summarize table to create the table.

The button Create summarized items appends the newly summarized items to the destination project selected in the first step.

You have the option to include or to ignore the item description modifiers or notes. If you include them and the same state from several items must be combined into a single new state, the notes will be appended, separated by a ";" and a new line character.

Categorical characters: The resulting set of character states is the union set of all items summarized in the new item. Thus if for character 2, item 1 has 2,1/2 and item 2 has 2,2/4, the resulting summarized item formed from item 1 and item 2 will have the states 2,1/2/4.

Text characters: data from several items are concatenated, separated by a ";" and a new line character. If two items have identical text entries, this text will not be appended a further time (even if other text was appended in between). Still, some manual editing may be necessary to obtain satisfactorily descriptions.

Numerical characters: The "summarized minimum" is the minimum of the minimum of the summarized items. The maximum is the maximum,

Example:

and with an additional item 4 added to item 1 to 3, this would look like

Note that in the summary of 1-3, item 3 does not influence the mean value. This is not optimal, but should the median of the range be included in the calculations?

In the second example of item 1-4, one can see that the minimum and maximum values are dropped, if they are already included in the normal range of at least one item.

Contents of Analysis    Main table of contents    Next