Skip to content

ANALYSIS VIEW🔗

The Analysis View (also called "dashboard", "post-processing dashboard", or "analysis dashboard") displays several plots in a grid. This Analysis View is connected to the currently opened/active simulation model.

Note

This article describes a new feature that is currently in a beta phase. To activate it, please set the corresponding toggle in the application settings. Please reload your browser tab after switching, so that the feature is activated.

The plots are arranged in a grid and you have two context-sensitive sidebars and One for the whole view.The Analysis View is only available in the Result layer.

Analysis View (Impact's post-processing dashboard)

If no plot is present or selected, you can use the right sidebar to select a Result (SIMULATIONS browser) and you can select a variable from the associated list of CALCULATED VALUES. On select a popover appears, where you can choose which kind of plot you want to add. Click on it will create/add a new plot in the grid or drag and drop variable in the grid.

Impact supports the following plot types with context-specific configurations and styling options:

If a plot is present, you can drag and drop further variables to this plot in order to compare variable progressions/values. Furthermore, you can drag and drop a variable to the x-axis to get a line chart or scatter plot over this variable.

To compare values/courses of different results for the same variable, In the SIMULATIONS browser switch to "multi-select" and select multiple results. One result must be the "primary result" (selectable via the options menu).

Set individual styling properties (like line style) or plot settings (like plot header), you can select a plot and use the offered options. Here, you can also use the VARIABLES browser, where you can choose to which variables the styling options are applied. Furthermore, you can remove variables from a plot there.

Finally, you can also rearrange the position of plots in the grid by moving them (click on the top bar of the plot), resizing the plots (drag the bottom left corner of the plot) or deleting plots (click on the cross in the top bar).

Tip

At the bottom of the Analysis view you will find a “Time and range slider”.
Use the playhead of the slider to select a point in time.
And use the range markers to zoom globally in all time-based charts.

Note

The Analysis View (all settings and changes) is automatically saved, so that you can access all your settings/configurations again later.
Only one dashboard per simulation model is supported.

Plots can have styling per plot or variable in a plot. The plot types and their styling/options are described below.
Also, general plot layout configurations and some general dashboard settings are supported.

Detailed descriptions for workflows on how to set up or configure plots can be found in the articles, linked at the bottom of this page.

Plot/Chart Types🔗

Modelon Impact supports the following plot types to analyze results:

  • Single-run time-domain (Dynamic simulation, variables calculated over time)
  • Single-run steady-state simulation (one value per variable)
  • Multi-run time-domain (Dynamic simulation for Parameter sweeps or Statistical analyses, variables calculated over time for each case)
  • Multi-run steady-state simulation (one value per variable and case)

Depending on the result kind you will find some context-related options.

Line Chart🔗

Line charts can represent the trajectories of variables ...

  • over time as y = f(time),
  • over another variable as y = f(x), or
  • as a family of curves over time/over another variable, y[i] = f(time)

Line chart of a multi-run result over a variable (y[i] = f(time))

For Line charts, you have the following styling options:

  • Line style: solid, dot, dash, dash-dot
  • Interpolation: linear, spline (useful if too less points were calculated and line interpolation looks too rough), step (for discrete jumps)
  • Type: lines, lines & markers, only markers
  • Fill: No fill, Fill to zero (... to the x-axis), Fill to next (... to the next graph)

Stacked Area Chart (Line Chart)🔗

Analyze which variable has a determined share of an overall result. For example, in the case of energy observations, it is recommended to check how much energy loss occurs in the system under investigation.

One way of doing this is the Stacked area chart.
Create a line chart with the corresponding variables, and use the Stack variables and Normalize options.

Stacked Area chart (Line Chart) with setting Normalize

Note

The order of variables affects the design of the stacked area chart. You can customize the order by using the VARIABLES browser.

Scatter Chart🔗

Scatter charts can represent the value tuples of variables ...

  • over another variable as y = f(x), which is useful for analyzing steady-state simulation, where the independent variable (x-axis) can be the modified parameter of a parameter sweep,
  • over time as y = f(time), which means you will see just one marker for the last value of time trajectory,
  • over all cases (multi-run result) as y[i] = f(case), where the independent x-axis represents just the case number (helpful e.g. in optimization/calibration analyses),

Scatter chart of a multi-run result over time, displayed over varied parameter

For Scatter charts you style the marker type: circle, star, or square.

For Scatter charts, you have also the option to plot over the case number (y[i] = f(case)), if the Scatter chart displays a multi-run result.
Select the plot and in panel EXPERIMENT in the right sidebar, you find the option X axis, where you can select "All cases" instead of "default".

Histogram charts🔗

Histograms are post-processing tools where you can evaluate data (trajectories) and represent the result as a bar chart. The histogram, a special form of bar chart, is used to display frequency distributions for statistical data. The characteristic values are entered on the x-axis and the frequencies on the y-axis.
The characteristic values on the x-axis are categorized/classified. In Impact this done automatically by just specifiy the total number of bars ("classes"). Their size is calculated based on the minimum and maximum number of the value set.

You can also specify the type of frequency normalization used for this histogram:

  • None (only count) (default): This absolute count corresponds to the number of occurrences (the number of data points inside the bars/"classes").
  • Probability / Percent: The span of each bar corresponds to the percentage/fraction of occurrences concerning the total number of sample points (the sum of all bar areas equals the total number of sample points).
  • Probability density: The area of each bar corresponds to the probability that a value will fall into the corresponding bar (the sum of all bar areas equals to 1).

Histogram comparing the probability density of 2 multi-run results

For Histograms, you can specify:

  • Number of bars: requires an integer number greater than zero
  • Normalization: see above

General Plot Layout Configuration🔗

You can set the following layout properties by selecting a plot:

  • Title: Give your plot a nice and unique title.
  • Legend: shows the option to view the legends in the bottom or right side of the plot
  • Show x-grid and Show y-grid: Show grids for the axes or not.
  • X precision and Y precision: Select between the default behavior (Auto) or a scientific notification with a specific number of digits.

General Settings🔗

The Analysis View supports some general settings, which are applied to the whole dashboard:

  • Show plot titles: Switching this off will enable more space for the plots. An already set title will not be deleted from the storage.

How to generate and work with plots🔗

The linked articles describe how to generate, configure, and work with plots: