Inspecting the traceability status, doing coverage analysis, calculating the impact of change, preparing meaningful management reports: That’s what the YT Analysis perspective is for. If instead you want create, modify, or delete links, please see the section on YT Editor .
The YT Analysis perspective has two major fields of activity:
The YT Analysis perspective consists of a couple of views that are, by default, arranged in a particular order and have particular sizes. You can change this arrangement at will, and you can even save your preferred arrangement as a new perspective or override the default settings of the YT Analysis perspective.
This section gives an overview of the YT Analysis perspective. You will find more detailed information in subsequent sections.
The dashboard provides answers to typical questions in traceability projects. For answers to not-so-typical questions, please see the section on customized queries.
Figure "The YT Analysis perspective; using the dashboard" exemplifies how this can look like:
The YT Analysis perspective; using the dashboard
Section "The YT Dashboard view" has more details on using the dashboard.
While the dashboard application answers common questions, it does not cover any specific needs of your particular project. However, this can be addressed with customized queries. Figure "Writing and running queries in the YT Analysis perspective" shows an example:
Writing and running queries in the YT Analysis perspective
Section "The YT Queries view and the YT Queries results view" has more details on working with customized queries.
To open the YT Analysis perspective, you have a couple of options:
In the YT Analysis perspective, you will find the YT Dashboard and the YT Queries tabs near the top-left corner of the YAKINDU Traceability window. At least this applies to the default settings of the YT Analysis perspective. You can rearrange the views making up the YT Analysis perspective as you like and save the result as a new perspective or override the YT Analysis perspective’s default settings.
The YT Dashboard and YT Queries view tabs
The YT Dashboard provides a graphical overview of commonly-used traceability metrics for the current project.
By choosing an element from the drop-down menu in the top left corner of the YT Dashboard view, you can choose which data to display in the diagram:
A counter right from each option in the drop-down menu indicated the number of links or artifacts found.
If you select a set of elements that is not empty, the dashboard displays a pie chart with an overview of these links or artifacts. It is grouped by link types resp. artifact types to give you an immediate visual impression of the element distribution in your project.
If there are no data available for the selected option, a notification will be shown in the center of the view.
Check the Show legend or Show labels checkbox near the top-right corner of the view in order to see which kind of elements each section of the pie chart represents. In figure "Pie chart and drop-down menu in the YT Dashboard view" , both have been checked.
Pie chart and drop-down menu in the YT Dashboard view
The pie chart labels are showing element information of the following format: "name: count (percentage%)".
If you hover your mouse pointer over a pie section or legend item, a tooltip displays the element information in the format explained above.
Pie chart section tooltip
The pie chart is interactive: You can click on any of its sections (or on any legend entry) to investigate the corresponding elements in more detail. A click on a pie chart section explodes it, highlights the corresponding legend entry (if the legend is visible), and displays the elements the pie chart section represents in the YT Dashboard Details view . Click again, and the actions are reversed.
Interactive pie chart sections in the YT Dashboard view
The YT Dashboard Details view is linked to the YT Dashboard. As the name suggests, it displays details on the elements selected in the YT Dashboard view. In YT Dashboard, you select one or more types of artifacts or links, and in YT Dashboard Details, you can see all the artifacts or links themselves.
The YT Dashboard Details view displays elements, i.e., links or artifacts, as a table. Each row represents one element and provides information on it. Artifact information are type, name, and origin tools. For a link, the view displays its type and the artifacts at the link’s ends.
If an element type is selected on the dashboard and the YT Dashboard Details view is closed, it will be opened. In any case, the YT Dashboard Details view will get the focus.
If no item is selected on YT Dashboard or the latter is closed, YT Dashboard Details provides either a clickable hint to open the dashboard or suggests to select a pie chart section.
Selecting an element in YT Dashboard Details displays its properties in the Properties view. If an artifact has been selected, the YT Overview view displays this artifact and its relationships to other artifacts. If a link has been selected, the YT Overview view displays its both link ends, i.e., the A and B artifacts, and their relationships to other artifacts.
The YT Dashboard Details view
YT Dashboard Details view provides a
copy to clipboard
button at the right-hand side of its title bar. Select the table rows you are interested in – pressing
all rows –, then click on the
copy to clipboard
button. The table heading and the selected rows are copied to the clipboard as plain text. Within each line of this text, table cells are separated by
Location of the copy to clipboard button
Instead of using the
copy to clipboard
button, you can press
[Ctrl+C] on the keyboard.
The YT Queries and YT Query Results views operate in concert and allow you to create, edit, and execute queries as well as to view their results.
Queries are stored in query files and displayed by the YT Queries view in a tree structure, with each query file at the top level, see the sample screenshot in the introductory section above.
The queries contained in a query file are shown as subordinate nodes of the respective file in the YT Queries view.
Query files must have the .query filename extension and can be stored anywhere in your workspace. A query file can contain multiple independent queries.
Initially, you don’t have any query files, so the YT Queries view will be empty.
Please note: Queries have no side effects and will never manipulate any data they are executed on.
You have several options to create a query file:
Both actions will open a file creation wizard asking for name and location of the file.
Creating a query file
After the query file has been created, YAKINDU Traceability adds it to the YT Queries view and opens it in an editor. In this editor, you can formulate your queries, using the YT query language.
You can add as many queries as you want to a query file, but it is advised to maintain different queries in different files or at least maintain a clear structure, depending on your queries' purpose and semantics.
To open a query file, you have several options:
The query file will be opened in the query editor.
You have the following options to execute a query:
Executing a query
After executing a query, the YT Query Results view shows the results. Please note that some queries may take considerable time to execute, dependent on their complexity and on the size of your data model. A progress bar will inform you about the status of the query execution.
Each query execution adds a new result tab to the YT Query Results view. That is, you can execute the same query several times without loosing any results of formerly executed queries. To differentiate query results from each other, each result tab gets an individual name assigned, consisting of the query name and a time stamp.
Inspecting query results
By default, query results are displayed in the YT Query Results view. However, it is possible to export them to a file, either to an Excel spreadsheet or to a CSV (Comma Separated Values) file. There are several ways to achieve this:
In any case, you will be prompted with a dialog asking for some parameters:
Exporting a query result
Please note: If the Excel document is already open in Excel, YAKINDU Traceability may not be able to write the file. In this case, a dialog will ask you to close the file in Excel and retry.
The YT query language enables dedicated analysis of and specific information from the whole trace model (i.e., the graph of artifacts and links, including unlinked artifacts, attributes of artifacts, and attributes of links) which has been created by YT based on the configuration and the information that has been extracted from the various data sources.
Queries are processed by the YT query engine in three steps:
A YT query consists of up to six elements which are listed in the table below and are described in more detail in the subsequent sections.
|Query meta data||
||Defines the display name of the query and an optional description.
In addition, a query may be declard to be
|the name is mandatory, markers private and validation and the description are optional)|
||Defines the algorithm a.k.a. query function which evaluates the trace model and returns a model set of result model set, i.e., the elements of the result set are not records, but objects, i.e., entities with attributes and relations. The expression may also contain parameters passed to the algorithm. This part can be roughly compared to the
||Defines the transformation of the result model set into a result set. This part can be roughly compared to the
||An expression restricting the rows of the result set||optional|
||Defines how the query results should be aggregated or projected.||optional|
||Defines the order of the rows of the query result.||optional|
Elements of a YT query are separated by a
. (dot), possibly with blank space left or right from the dot. An exception are
query meta data and
data source: These query elements are separated by blank space. This is illustrated by the following example which utilizes all language elements.
query "name" description "description of the query" source(someQueryFunction(parameter1, parameter2)) .collect(attribute1 as A1, @sum(attribute2) as A2) .groupBy(attribute1) .orderBy(A2)
The expression specified in a
source(expression) clause links to an algorithm which processes the trace model and returns a set of
result models. The algorithm itself is a so-called
query function implemented in Java. The
expression can pass parameters to the query function. YT does not restrict the types of the model elements of the result model set in any way.
A crucial characteristic of a query function is its return type. While YT does not restrict the element types of a result model set, all elements in the result model set of a given query are of the query function’s return type.
As an example, YT provides the following query function:
Iterable<TVMDArtifact> allArtifacts(TVMCArtifactType type)
The allArtifacts function is one of YT’s built-in query functions. The definition above shows that allArtifacts takes a single parameter of type TVMCArtifactType, which represents any artifact type. The result model set returned by allArtifacts contains all artifacts of the specified artifact type. It comes in the form of an Iterable of TVMDArtifact, i.e., as a sequence of zero or more TVMDArtifact objects, each of them representing an artifact of the specified type.
Consequently, a YT query using the
allArtifacts function in a
source clause, as in, e.g.,
source(allArtifacts('TestCase')), populates the result model set with matching
TVMDArtifact elements, in this case with elements of type
YT query clauses right from the
source(allArtifacts(type)) clause operate on these elements. Like all
TestCase has a
name and a
version. So a query like
query "simple example" source(allArtifacts('TestCase').collect(name as n, version as v)
creates a result set with columns n and v containing the name and the version of the TestCase artifacts.
Generally speaking, query clauses right from the
source clause like
where, etc. are based on the element type of the result model set and consequently on the query function’s return type. If, for example, you replace the query function expression
allArtifacts('TestCase') by a different query function, say
foobar('TestCase') with a different return type, the whole query might become invalid.
Apparently, being versed in the YT query language requires knowledge not only of the query language syntax, but also of available query functions, their return types, the specifics of these types, and the data model as a whole.
Please find a complete overview of the built-in query functions and their return types in section YT built-in query functions.
Since YT recognizes query functions dynamically during runtime, it is easy to add custom query functions to YT and in this way extend the YT query language with evaluations that are specific to your development process.
collect clause defines which attributes of the model elements returned by the query functions are taken over as a result set row of the YT query. A
collect clause defines rules for the relevant attributes column by column, the rules being separated by comma. Each individual attribute is defined by navigating the attributes of an element of the result model set. The syntax is a dot-separated expression defining the access to attributes and getter functions, similar to, e.g., the C# programming language.
As an example, the query function tracesFromTo returns a set of traces.
So, a query like:
query "Requirements covered by test cases" source(tracesFromTo('Requirement', 'TestCase')) .collect(Start.DataSource.Identifier as FileName, Start.Name as ID)
returns a result set containing FileName and ID of Requirements which are (not necessary directly) linked to TestCases. A Requirement which has n TestCases assigned is listed n times in the result set.
collect clause, you can not only access attributes, but also apply aggregator functions. As aggregator functions are making sense with an aggregation only, a
groupBy clause is mandatory if using such functions. Available aggregator functions are:
||Calculates the average, maximum, minimum, or sum, respectively, in an aggregation. The specified expression must evaluate to a numeric value.|
||Counts the number of elements. The expression is irrelevant for counting the elements, but it must be specified and valid. For example,
||Returns the first resp. last element of an aggregation.|
||Joins the elements of the expression to a string by using the separator string passed as the second argument or '
@sum only work on numbers and strings that can be converted to numbers (both variants can be mixed arbitrarily). Their result is always a number. In case of
@avg, the result will be rounded to 5 digits in the fraction. For example,
@avg for (2.3 ,7, „1e-12”) can be computed, but „1e-12” is too small to affect the result, because of the limit of digits in the fraction.
@min work on arbitrary parameters, but will work differently depending on the kinds of parameters. If all parameters are numbers or strings that are convertible to numbers, they will be compared numerically. If at least one of the parameters is not convertible to a number, the string representations for all parameters are compared alphanumerically. For example, the result of
@min is 10 for („x”, 2, 10), while it is 2 for (2, 10). The result of
@max for („xa”, „xb”, „X”, 1) is „xa”, because lower case letters are alphanumerically larger than upper case letters.
Example: The query
query "Requirements covered by test cases" source(tracesFromTo('Requirement', 'TestCase' )) .collect(Start.Name as ID, @count(End) as Coverage) .groupBy(Start.Name)
returns a result set containing the IDs of Requirements and the number of (not necessary directly) linked TestCases.
Values of a result set are matched against the boolean expression specified by a
where clause. Within a
where clause, boolean expressions can be combined using
and logic. Grouping them by parenthesis
(…) is supported. The available operators are as follows, in the order of highest to lowest precedence:
|Primary expression||attribute||Attribute evaluation, see Collect clause|
||Logical NOT, negative number, positive number|
||Multiply, divide, modulo|
||Larger than or equal, smaller than or equal, larger than, smaller than|
|Logical AND operator||
||Combines two boolean expressions by a logical AND. The resulting expression evaluates to true if the combined conditions both evaluate to true, else false.|
|Logical OR operator||
||Combines two boolean expressions by a logical OR. The resulting expression evaluates to true if at least one of the combined conditions evaluates to true, else false.|
Example: The query
query "Requirements covered by manual test cases" source(tracesFromTo('Requirement', 'TestCase')) .collect(Start.Name as Id, @count(End) as Coverage) .where(Start.AttributeValue("Status") == "Verified" && End.AttributeValue("Execution Kind") == "Manual") .groupBy(Start.Name) .orderBy(Coverage)
finds all verified Requirements with their respective number of linked manual TestCases. The requirements are grouped by the name of their sources. Within each group they are ordered by their coverage, i.e., by the number of test cases assigned.
Every so often you are more interested in certain higher-level characteristics of a result set than in the individual members of a query result. A basic example would be to return the total number of requirements that are associated with at least one test case and the total number of those that don’t.
This requires to first group the result set elements by an attribute like, say, hasTestCases, and then count the members of both groups using the @count aggregator function. Such a group is called an aggregation.
Another example is shown near the end of the
Collect clause section above. The query function
tracesFromTo('Requirement', 'TestCase') returns all traces from requirements to directly or indirectly associated test cases. Requirements having multiple test cases appear multiple times in the resulting traces. Now essentially two things happen:
groupBy(Start.Name)puts all trace elements with identical requirement names into a group of their own, respectively. That is, if a requirement has, say, five test cases, then five of the trace elements returned by the query function will contain that requirement as their start attribute. These trace elements will be grouped into an aggregate of their own, just like other trace elements with identical requirement names.
collect(Start.Name as Id, @count(End) as Coverage)clause, the
@count(End)aggregator function is particularly interesting, because it counts the elements in each aggregation. The sample
collectclause creates a result set with two columns: Column Id holds the requirement names, and column Coverage holds the number of test cases for each of those requirements.
Aggregator functions work on aggregations and thus
groupBy clause to be present in the YT query.
It is possible to group a result set by more than just one criterion. Multiple feature selections are specified as a comma-separated list, e.g.,
groupBy(A, B, C).
orderBy clause is used to sort the result set by one or more columns. An example is given in section
YT comes with a lot of built-in query functions. As foreshadowed in section YT query language elements, these functions do not return plain data records, but a set of result models, e.g., models containing instances of artifacts and links. The result models returned by the built-in YT query functions are all based on the same meta-model. The returned models are complete; if, e.g., a function returns a set of links, it is possible to navigate from such a link first to the artifact at end A and then from there to the artifact type.
|Function||Description||Parameters||Return type, i.e., set of …||Example|
|allArtifacts||Finds all artifacts of a given type or category.||Type: ArtifactType or Category||Artifact||
Returns all names of all test cases.
|allArtifactTypes||Lists all artifact types specified in the YT configuration.||–||ArtifactType||
Returns all names of all artifact types.
|allLinkTypes||Lists all link types specified in the YT configuration.||–||LinkType||
Returns all names of all link types.
|allTraceLinks||Lists all links for a given type or classification||Type: LinkType or Classification||Link||
Returns both ends of all links between requirements and tests.
|artifactsWithoutTraceFromTo||Finds all artifacts of one type or category that have no traces to any artifact of a second type or category. When a category is used as a parameter, the paths are computed for every combination of an artifact type of that category with the artifact type(s) defined by the other parameter.||TypeA:
TypeB: ArtifactType or Category
Returns all customer requirements without any assigned test results. „Assigned” in this case means „reachable via a chain of links”, e.g., from Customer Requirement to System Requirement to System Test Case to Test Result.
|intersection||Creates the intersection of two given query results. Rows of the results are identical when all columns and values are identical.|| TypeA:
Returns a query result that contains the „artifact” column for all elements contained in the query results of both q1 and q2.
|join||Joins two query results (i.e., from q).|| Query1:
Returns the „Name” of „q1” and all "End"s for these artifacts in „q2”.
|linkedArtifacts||Finds all artifacts of a given type that have at least one link.||Type: ArtifactType||Artifact||
Returns all names of all linked test cases.
|notLinkedArtifacts||Finds all artifacts of a given type that have no link.||Type: ArtifactType||Artifact||
Returns all names of all test cases which are not linked.
|q||Get the result of one query as input for another one like join or union.||Query: Query||QueryResult||
Returns the column „Name” from „AnotherQuery”.
|reduceBy||Subtracts the query result given as second parameter from the query result given as first parameter. Rows of the results are identical when all columns and values are identical.|| TypeA:
Returns a query result that contains the „artifact” column for all elements contained in the query result of q1, but not in that of q2.
|tracesFromTo||Finds all traces (i.e., chained trace links) between artifacts of two artifact types/categories. When a category is used as a parameter, the paths are computed for every combination of an artifact type of that category with the artifact type(s) defined by the other parameter.|| TypeA:
TypeB: ArtifactType or Category
Returns all customer requirements with at least one assigned test result and the number of these assigned test results. „Assigned” in this case is defined as above.
|traceMatrixFor||Calculates all possible traces between two artifact types, while also considering incomplete, or interrupted traces. In the collect clause, you can access the attributes of artifacts in the trace. To access all artifacts, use ArtifactsByType(type). You can also access the first, last, or Nth artifact in the trace with FirstArtifactByType(type), LastArtifactByType(type), and ArtifactNByType(type, index).|| TypeA:
Assuming that the trace model looks as follows: „Customer Requirement” → „Software Requirement” → „Implementation” → „Test Case” → „Test Result”. The result will contain all distinguished traces between each and every artifact in the chain. Also unlinked artifacts are included. The last column contains the values of the custom attribute „result” for each „Test Result” artifact in the trace.
|union||Creates the union of two given query results, i.e., the set of elements that are contained in either of the results. Duplicates are discarded, so when there are several identical elements in the results, only one is retained. Elements are considered identical, if all columns and values in their respective rows are identical. Elements are not guaranteed to be in any particular order.||TypeA:
Returns a query result that contains the „artifact” column for all elements in the query results of q1 or q2. Each element is contained only once.
|unionAll||Combines two given query results by appending the elements of the second result to the elements of the first result. All duplicates are retained, and they remain in their original order.||TypeA:
Returns a query result that contains the „artifact” column for all elements in the query results of q1 or q2. Duplicates are not removed.
For the results of the built-in YT query functions the most important domain objects and their relations are illustrated and described below.
DataSource: Represents the origin from which artifacts or links have been extracted. For the Microsoft Word Adapter, for example, a DataSource represents the Microsoft Word Document. For the IBM DOORS adapter, the DataSource represents the DOORS module.
Link: A single trace link between two artifacts.
collect(… AttributeValue("my custom Attribute") as MyCustomAttribute).
Artifact: Represents a traceable artifact that may or may not have links.
collect(… AttributeValue("my custom Attribute") as MyCustomAttribute ).
Trace: Represents a chain of linked artifacts, i.e., the n th artifact in the chain has a direct link from the artifact at position n-1 and a direct link to the artifact at postion n+1.
LinkType: Reflects an artifact type specified in the YT configuration.
Note: From a technical point of view, the above-listed attributes of a link type are not shallow attributes, but attributes of related objects, e.g., the attribute classification.name is the name of the classification the link type is related to. Nevertheless, from a domain point of view, „attributes of related objects” can be viewed as attributes of the link type.
ArtifactType: Reflects an artifact type specified in the YT configuration.
The report wizard can be accessed by navigating to the menu entry
Traceability → Reporting → Create report
The default report shows simple traceability statistics only. Further reports (e.g., coverage analysis report) have to be implemented by the end users. The reporting is based on Eclipse BIRT (Business Intelligence and Reporting Tools). Reports can be designed with the BIRT WYSIWYG Designer. YT comes with data access drivers for BIRT to provide convenient access to traceability data.
You have the option to select one or several validations to be executed before the report creation. This is to ensure that the report is created based on valid data and will contain reliable results.
In the above example a coverage report is designed that shows User Requirements without System Requirements. Coverage analyzes are calculated by joining and filtering corresponding data sets.
During the design of reports often all possible artifacts of a specific type are required. These sets are called candidates. The figure below depicts the meaning of candidates.
YT can export the configuration and the data as a graph. The export is a GraphViz dot-file that can be rendered with GraphViz. The files are generated by navigating to the menu entry
Traceability → Export → Export Configuration
Traceability → Export → Export Data
Below are sample exports of configuration and data.
GraphViz can be downloaded from http://www.graphviz.org/.