11. Pages and Visualizations in Detail

In this chapter, we will take you through the steps involved in performing a detailed quality evaluation in MQC. This will give you a good impression of the functional details of MQC. At the same time, it helps you get accustomed to the tool in a more detailed way than in Quick Start Guide where a quick introduction to the basic functionality of MQC is given.

Building on this and taking into account the pages for a project, conclusions about the development of software projects are described. Based on these conclusions MQC shows possible reasons for bad quality and helps you deducing actions to carry out for improving quality.

11.1. The MQC Pages Concept

MQC structures all the important information of your project and shows it on different pages. Mainly, there are two kinds of pages, showing either data or quality, containing trend and status visualizations respectively, i.e. Quality Trend and Quality Status, as well as Data Trend and Data Status.

The Quality Trend page will give an overview of the quality of your project, whereas the Quality Status, Quality Sunburst and Quality Heatmap page provide an in-depth analysis of Artifacts, i.e. an overview of the quality of each Artifact.

The Data Status page will give you an overview of the data of your project. By means of Data Details functionality (Data Details) you can easily navigate from quality pages, such as Quality Status or Quality Sunburst, to the data pages to track the source of a measurement. In other words, going one level down at the data source level to Data Trend where you can see your raw data’s trend and get an overview of your related base measures.

On all status pages the main visualization always shows one revision with the possibility to switch revisions from the dropdown menu above. By default, all revisions means all revisions with imported data. For example, if revision granularity is set to Days, but not every day something is imported, days without imported data are not shown in the visualizations.

The general concept of all pages is that all the visualizations are interactive: If you click into any of the visualizations, the other visualizations will react and then show information related to what you have marked. You can do the major selections on the left and top sections of the visualization window and the effect can be seen in the main/bottom right visualization.

Marking is cumulative. The marked elements of the first chart will not reset if you mark something in the next chart later. The effect is a combination of marking done in both the charts. To go back to normal state click on the chart where there are no elements.

Based on the Project Structure you have imported, the right side panel will display options for filtering the data.

All MQC pages can be easily added or removed using the Pages dialog of the left-hand side panel (see Manage Pages).

In the following subsections, the main pages are described in detail.

11.2. Dashboard

The Dashboard page gives a first overview of the current status of the most important aspects of the project at a glance. It can be used as a starting point for quality evaluation and to explore the details of quality deficits in your products. MQC provides a default dashboard page, as explained below, which can also be customized based on your requirements (see Dashboard Customization).

Each visualization in the default dashboard shows a particular aspect of quality assessment in a compact way, leaving out any details that can be found in the main visualizations of the MQC standard pages. For that purpose, MQC provides a link to the corresponding detail page in the top-right corner of each visualization.

If not visible, the Dashboard page may be added using the Pages dialog of the configuration panel (see Manage Pages).

Marking does not apply to the Dashboard page. But instead, filtering the data via the filter panel at the right-hand side may be used to concentrate on specific aspects.

The Default Dashboard page consists of the following visualizations:

  • Project Information (top-left):

    The Project Information provides some general information regarding the project. Here you will find the number of days left until the end of the project and until the next milestone, as well as the current revision.

    In addition to the overall quality and availability values and the number of artifacts, the number of data points is shown, which means the size of the data loaded into MQC and used for quality calculation.

  • Quality Trend (top-right):

    The Quality Trend diagram shows 3 aspects of the overall quality of the project and how it evolves over time. Vertical lines display the project milestones.

    • Absolute Quality: This overall project quality is calculated using all data also including missing data.

    • Available Quality: This overall project quality is calculated using the available data only, excluding the missing data. This directly shows the quality of the things already done.

    • Relative Quality: The third quality trend line shows the overall project quality in relation to targets configured for particular milestones (see Target Values). An absolute current quality of only 50 % is quite bad, but if the plan was to achieve 50 % at that point in time, it is excellent (100%).

Note

If no specific targets were imported, Absolute Quality and Relative Quality are showing the same trend lines.

  • Actions (middle-left):

    The Actions visualization shows the actual tasks to improve the current quality. This is based on most recently loaded data. It provides a sorted list of actions per artifact as well as a distribution of actions according to priority groups.

    For details on actions see Actions.

  • Quality Status (middle):

    The Quality Status shows the 4 artifacts with worst quality as well as the quality distribution according to the defined quality bins in general. As per default this means how many quality values are

    • good (over 80 % quality)

    • acceptable (20 % - 80 % quality)

    • bad (under 20 % quality)

    • missing (i.e. quality could not be calculated because of missing data).

  • Data Availability (middle-right):

    The Data Availability shows the 4 artifacts with the worst availability status together with the distribution of availability of the loaded, propagated (from previous revisions), and still missing data in general.

  • Quality Breakdown (bottom-left):

    The Quality Breakdown visualization shows a condensed Sunburst chart (see Quality Sunburst) just containing the overall quality (centre) and the quality characteristics (outer ring) as defined in the quality model (see Quality Model Structure).

    By this, you are able to see the project quality for specific aspects like “Correctness”, “Completeness” or “Consistency” at a glance. You may also define different characteristics within your quality model to focus on other facts, e.g. “Functionality” or “Test Progress”. For more information see Quality Model Configuration.

  • Quality Diff By Artifact (bottom-midst):

    This chart provides a short overview how quality has been changed since the previous revision. Per artifact, it is shown if quality has been increased or decreased.

  • Availability Diff By Artifact (bottom-right):

    This chart provides a short overview how availability has been changed since the previous revision. Per artifact, it is shown if availability has been increased or decreased.

11.3. Data pages

11.3.1. Data Trend

By means of the Data Trend page, you can track the source of a measurement, going one level down compared to the Data Status page, i.e. on data source level. You can see your raw data’s progression and processing getting an overview of your derived and related base measures.

It consists of the following selecting windows:

  • Projects (on the top left): Select your project as defined in your project structure. Notice, that you see an availability trend inside the tile. A gradient availability coloring is used that depends on the availability of all measure values for all artifacts shown for the currently selected revision.

  • Artifacts (Bottom left): Select the artifacts of your choice. By default, all are selected. A gradient availability coloring is used that depends on the availability of the measure values for a single artifact for the currently selected revision.

  • Measures (next to Artifacts): The count of the measures is with respect to all artifacts and for the currently selected revision. It uses a gradient availability coloring.

  • Availability (Bin) Distribution: This chart shows the trend of the data availability as bars per revision. For detailed information about distribution charts see Distribution Charts.

The Measure Trend visualization shows per Artifact for each measure value (base measures and derived measures) a trend over all revisions. If a certain revision is selected, the data value for this specific revision is shown only. It is possible to only show the trends for a single or for selected artifacts as well as for selected measures. The Measure Trend visualization uses a dedicated color per trend line for each measure value.

11.3.2. Data Status

The Data Status page offers the following selecting windows:

  • Projects (on the top left): Select your project as defined in your project structure. Notice, that you see an availability trend inside the tile. A gradient availability coloring is used that depends on the availability of all measure values for all artifacts shown for the currently selected revision.

  • Artifacts (Bottom left): Select the artifacts of your choice. By default, all are selected. A gradient availability coloring is used that depends on the availability of the measure values for a single artifact for the currently selected revision.

  • Data Sources (next to Artifacts): The availability of the measures per data source (and measurement, e.g. “TPT >> MiL” and “TPT >> SiL”) is with respect to all artifacts and for the currently selected revision. It uses a gradient availability coloring.

  • Availability (Bin) Distribution: This chart shows the trend of the data availability as bars per revision. For detailed information about distribution charts see Distribution Charts.

The Data Status chart is the reacting window to the selections (intersections) made by the marking in the selecting windows mentioned above. It also uses a categorical availability coloring.

Note

If data from one data source for multiple measurements was imported, marking one of the data source KPI tiles will select all “DataSource >> Measurement” tiles. A second click on a specific tile selects a particular data source and measurement.

11.4. Quality pages

Quality pages will give you an overview of the quality of your project. They have the same structure and selecting windows:

  • Project (on the top left): See the overall quality for your project. You can also reset the marking for the entire page at any point by clicking on this tile. Notice, that you see a quality trend over all revisions inside the tile. A gradient quality coloring is used that depends on the overall quality for all artifacts, shown for the currently selected revision.

  • Artifact Structure (left): Select the Artifacts of your choice. By default, all are selected. The values within the tiles always show the currently marked revision, which is the last available revision if none has been marked. A gradient coloring visually indicates the quality for each artifact. By using the dropdown or the arrow-buttons a different grouping inside the configured artifact structure can be selected. If a structure-element inside a hierarchy is choosen, the tiles offer the possibility to expand up until the artifact-level of the current hierarchy. The search field offers the possibility to temporary limit the shown tiles, while the sorting-dropdown allows for a different order of the tiles.

  • Quality Structure (next to Artifact Structure): Similar to the Artifact Structure this visualization works pracically the same, just for Quality Properties and the configured quality structure.

    ../_images/MQC_QualityStructureKPI.png

    Figure 11.1 Quality structure with selected Characteristics-Level and expandable tiles.

  • Quality (Bin) Distribution: This chart shows the trend of the project quality as bars per revision. For detailed information about distribution charts see Distribution Charts.

  • Main Visualization (Bottom right): This visualization is a page specific chart presenting the quality values in form of matrix, heatmap, sunburst, line or bar chart. Depending on the visualization different dropdowns are provided to select a specific revision and/or to select a grouping in the artifact structure and/or the quality structure.

On Quality Page you can choose which Scope of quality assessment to be shown on the visualizations from the dropdown in the menu panel on the left-hand side.

../_images/MQC_Select_Quality_Scope_Dropdown.png

Figure 11.2 Dropdown menu for selecting scope of quality assessment.

  • Absolute: This is the project quality. For quality assessment, all the expected and configured data is used for the calculations and shown in the visualizations. In case that quality cannot be calculated or the data is currently missing (i.e not yet executed tasks), a value of 0 (missing = bad) is used in quality aggregations. This is the default selected scope.

  • Available: In quality aggregations only the provided data (i.e already done or started tasks) is used for the calculations and shown in visualizations. Missing data is ignored and not shown.

  • Relative (Target): This option is available only when you had defined Target Values for the project. The quality of the project is based on the targets configured for particular milestones (see Target Values), which means you can compare the progress of your project with the planned progress defined in Target Values. For example current quality of only 50 % is quite bad, but if the plan was to achieve 50 % at that point in time, it is excellent (100%).

You can now start identifying the reason for bad quality. The general concept of identifying issues is to click on (or hover over) red quality bins or tiles. When clicking on red, data that corresponds to green and yellow bins is excluded from the visualizations to receive a first impression of the reason for bad quality, concretely the quality properties that have failed.

11.4.1. Quality Status

The Quality Status page provides an in-depth analysis of the quality of Artifacts. Therefore, it will give an overview of the quality properties of each Artifact.

The reacting window is a heatmap visualization plotted for each Artifact against each Quality Property. Each Quality Property tile has a gradient quality coloring, which represents the quality value for the particular property. Based on the marking from the selecting windows, the main visualization is reduced to the selected items.

11.4.2. Quality Sunburst

On the he Quality Sunburst page, the main visualization shows the sunburst visualization where the outer ring consists of all quality properties defined by the quality model (see section Quality Properties).

11.4.3. Quality Heatmap

On the Quality Heatmap page, the main visualization shows a more detailed heat map visualization, where quality properties and artifacts are visualized with their aggregation and respective “size” (weight).

11.4.4. Quality Trend

On the Quality Trend page, Quality Trend by Artifact and Quality Trend by Property are the reacting windows. These line chart visualizations are updated based on your previous selections.

The trend visualizations use a categorical coloring, so each line representing an Artifact resp. a quality property uses a dedicated color.

Note

Values between -10 and 10 on Dashboard, Quality and Data pages are shown with one decimal point.

11.5. Distribution Charts

A distribution chart is an integral part of each MQC page (always top right). It shows the trend of either quality or data availability as bars per revision (see Figure 11.3).

../_images/MQC_Visualizations_QualityBinDistribution.png

Figure 11.3 Quality Bin Distribution chart showing the portions of each quality category per revision.

11.5.1. Display Mode

As per default, a distribution chart either shows Quality Bins By Revisions (Quality pages) or Availability Bins By Revisions (Data pages).

  • Quality Bins: The computed quality property measures within a revision are grouped according to the configured quality bins (see Quality Model and Quality Bin Configuration). This means, depending on its value a measure is assigned to a specific category. For example, a quality measure value of 95 % belongs to the Bin category Good, when this category covers a range of ]80%, 100%].

    The default quality bins used by MQC are Good, Acceptable and Bad.

  • Availability Bins: The data within a revision is grouped according to the availability. The availability bins used by MQC are Available (data is imported in the current revision), Propagated (data is taken from a previous revision, see Data Propagation) and Missing (no data imported so far).

For the bin distribution a categorical coloring is used (see Color Schemes).

Additionally, as shown in Figure 11.3 the bin distribution display mode contains a trendline, which is either the overall quality (Quality pages) or the overall availability (Data pages).

The display mode of a distribution chart can be changed by clicking on the Settings icon in the top-right corner of the visualization (see Figure 11.4).

../_images/MQC_Visualizations_QualityBinDistributionSettingsIcon.png

Figure 11.4 Switch the display mode between quality/availability bin distribution and overall quality/availability via the Settings icon.

Figure 11.5 shows the overall quality as bars per revision, using a gradient coloring (see Color Schemes).

../_images/MQC_Visualizations_OverallQualityDistribution.png

Figure 11.5 Quality Distribution chart showing the overall quality as bars per revision.

11.5.2. Milestones

Project milestones (see Milestones) are shown as vertical lines including the corresponding milestone name between the revisions (see Figure 11.6).

../_images/MQC_Visualizations_QualityBinDistributionMilestones.png

Figure 11.6 Project milestones shown as vertical lines between revisions.

This eases the temporal recognition of the quality or availability distribution over time.

All revisions before a milestone can be collapsed. Collapsing revisions provides a streamlined trend (quality or availability) especially for long running projects with many revisions. To collapse the revisions for a milestone, click on the left-hand side icon on top of that milestone period (see Figure 11.7).

../_images/MQC_Visualizations_QualityBinDistributionCollapseIcon.png

Figure 11.7 Collapse all revisions before a milestone by clicking on the left-hand side icon on top of the milestone period.

When collapsing a milestone period, the distibution chart shows the last revision of that milestone.

Note

Marking the last revision of a collapsed milestone, selects all revisions within that milestone period.

If not all revisions of a collapsed milestone are marked, this is depicted by a lighter coloring.

Additionally, it is possible to focus on a particular milestone period by clicking on the right-hand side icon on top of that milestone period (see Figure 11.8).

../_images/MQC_Visualizations_QualityBinDistributionFocusIcon.png

Figure 11.8 Focus on a particular milestone by clicking on the right-hand side icon on top of the milestone period.

When focusing on a milestone, the distribution visualization only shows the revisions belonging to that particular milestone period.

11.6. Color Schemes

As per default on Quality Pages (i.e. Quality Status page, Quality Sunburst page and Quality Trend page), MQC uses the traffic light color scheme:

Categorical Coloring

good

Good

All quality properties with an evaluated quality between 80% - 100%

accp

Acceptable

All quality properties with an evaluated quality between 20% - 80%

bad

Bad

All quality properties with an evaluated quality between 0% - 20%

miss

Missing

Quality cannot be evaluated because of missing quality measure values

The quality categories (bins) and colors for the categories may be adapted, see Quality Model and Quality Bin Configuration.

Additionally, as shown in Figure 11.9 MQC uses a gradient quality coloring for certain elements that have a computed or aggregated quality value assigned (e.g. for projects, artifacts and quality properties).

../_images/MQC_Details_QualityColoring-gradient.png

Figure 11.9 Gradient Quality Coloring used for elements that have a computed or aggregated quality value assigned

On Availability Pages (i.e. Data Status page and Data Trend page) MQC uses a blue coloring scheme. This means you can easily distinguish between available data in blue and missing data in grey.

Categorical Coloring

avai

Available

Measure data is available for a certain revision

prop

Propagated

Measure was missing, but could be propagated from previous revision

miss

Missing

Measure data is missing for a certain revision

Additionally, as shown in Figure 11.10 MQC uses a gradient availability coloring for certain elements that have an aggregated availability (e.g. for projects and artifacts).

../_images/MQC_Details_AvailabilityColoring-gradient.png

Figure 11.10 Gradient Availability Coloring used for elements that have an aggregated availability

11.7. Structure Levels

Projects of different sizes need different levels of abstraction, e.g. a number of artifacts can be grouped to a certain software component whereas the project itself then consists of multiple components. A user may not only be interested in checking the status of single artifacts. Rather, the user could like to observe the quality on higher levels in addition.

The structure levels give the option to view the same data in a more aggregated form on higher levels and to filter or mark the desired details.

MQC allows to set up structures for different items:

To define an ‘Artifact Structure’, a project structure configuration has to be created (see Project Structure Configuration).

The ‘Quality Structure’ can be customized with a quality model configuration (see Quality Model Configuration).

The ‘Measure Structure’ is defined by the DataSource Adapters, that import the data.

Figure 11.11 shows the different levels of a quality model visualized using a Sunburst chart.

../_images/MQC_StructureLevel_Sunburst.png

Figure 11.11 Quality Structure visualized in the Sunburst main visualization

These defined and/or default structures for Artifacts, Quality Properties and Data Measures are visible in the filter panel and different visualizations (e.g. the Sunburst).

../_images/MQC_StructureLevel_ArtifactStructureFilterPanel.png

Figure 11.12 Artifact Structure in the filter panel

In addition, structure levels are also shown in KPI visualizations. These visualizations provide a level selection to modify the visible levels. It is also possible to expand selected (or all) levels to get a tree view. Click on [+] or [-] to expand or collapse a level.

../_images/MQC_StructureLevel_ArtifactStructureKPI.png

Figure 11.13 Artifact Structure in the KPI visualization

11.7.1. Structure level selection in KPIs

Each of the KPI visualizations for artifacts, quality and measures allow to change the visible level of the structure within. The level can be changed by selecting the desired level with the dropdown or using the buttons on the left and right side of the dropdown.

../_images/MQC_StructureLevel_KPIQualityLevelSelection.png

Figure 11.14 Level selection dropdown of the ‘Quality Structure’ KPI visualization

If there are more than one hierarchies configured, switching the level also means switching to the hierarchy the level belongs to.

Besides the dropdown, the level selection contains additional buttons to provide a faster way to navigate the levels and switch between them.

../_images/MQC_StructureLevel_KPIButtons.png

Figure 11.15 Quick Switch buttons on the ‘Artifact Structure’ KPI visualization

11.7.2. Structure level selection in the main visualization

Depending on the main visualization level, selections for the artifact structure, quality structure and/or measure structure are available.

The visualizations aggregate the values based on defined weights to show the quality or data to reflect the selected level.

../_images/MQC_StructureLevel_KPIQualitySubcharacteristics.png

Figure 11.16 Quality Matrix main visualization with ‘Subcharacteristics’ level selected

Note

Typically, the weighted average of all lower element values, e.g. quality properties, is used for aggregation when selecting a higher level.

This does not apply to the Availability Matrix shown on the Data Availability page. Here, the color of the worst contained element result is shown on higher levels. For example, if at least one measure for an artifact is missing, the whole data source is shown as missing, too.

By default the visualizations on the quality pages are configured to show the lowest level of the structure while the ‘Measure Structure’ on the data pages shows the Measurements level instead of the Measures level to give an better overview.

11.7.3. Search and sorting

Below the level selection there is also a search and sorting feature to customize the visualization further to your needs.

../_images/MQC_StructureLevel_KPIFeatures.png

Figure 11.17 Toolbar of the ‘Artifact Structure’ and ‘Quality Structure’ KPI visualizations

Any search term is applied to the current level and the search result is directly shown in the KPI visualization while typing by removing all not matching elements. For details how to easily mark multiple search results, see KPI marking.

The Sorting dropdown allows to adapt the order of the KPIs, i.e. sort by name or by value in both cases ascending or descending.

11.8. Marking

The purpose of marking is to restrict the shown data to the interesting parts.

The following should be considered while using marking:

  • When marking one or multiple elements of a visualization, the data that corresponds to other, not marked, elements are excluded from all visualizations.

    As seen in Figure 11.19, by marking data in one visualization except the main visualization, only data representing this marking selection is shown in each of the other visualizations.

../_images/Marking_DataStatus_Full.png

Figure 11.18 Data Status page without any marking

../_images/Marking_DataStatus_Artifact_Marked.png

Figure 11.19 Using artifact marking on Data Status: All other visualizations show only data related to the selected artifact (GlobalPosition)

  • Please note that marking works cumulative. That means further marking without resetting will reduce the resulting set of data even more. To mark multiple elements in the same visualization, press and hold Ctrl before clicking on another element. It is also possible to draw a rectangle enclosing the interesting parts by clicking and holding down the mouse button. (e.g. in Distribution charts).

  • Resetting the markings can be achieved by clicking on the Project KPI tile.

  • On Status pages (Data Availability, Data Status, Quality Status, Quality Sunburst, Quality Heatmap) a revision is marked with just one click in the distribution visualization. With a second click a selected bin (e.g. a quality bin) is marked.

  • On trend pages (Data Trend, Quality Trend) quality bins can be marked for all revisions with one click.

../_images/Marking_QualityTrend_Bin.png

Figure 11.20 On the Quality Trend page the first click on an quality bin will select the respective quality bin for all revisions

  • Artifact and revision marking on a page will result in the same marking on other quality, data, and action pages.

../_images/Marking_QualityTrend_After_Datamarking.png

Figure 11.21 Quality Status page after marking GlobalPosition on data pages (see Figure 11.19)

  • After using the search feature of a KPI visualization the marking of the results can be achieved with clicking on the ‘Mark all visible’ button.

    Holding ctrl to click the button marks the visible items in addition to an already existing marking. Holding ctrl and click again removes the selected element from the list of marked items.

../_images/Marking_QualitySearched.png

Figure 11.22 ‘Quality Structure’ searched for ‘coverage’