Paraview Basics
Paraview can read a number of different file formats, including the Exodus format, which is the primary output format for MOOSE.
Paraview Workspace
The Paraview workspace consists of the following basic areas:
Header rows: These contain a number of different buttons and dropdowns for various operations such as opening files, playing transients, controlling views, and creating filters.
Pipeline browser (Located on the left, directly under the header rows): This shows the loaded files and the pipeline of filters applied to those files.
Properties Pane (Located underneath the pipeline browser): This shows a number of different configurable properties that apply to the selected pipeline element.
View pane (Located in the main, large area): This contains the views for the output, which could be renderings or line plots, for example.
Opening a File
To open a file, do the following:
File->OpenSelect the file and click "OK".
In the properties pane, select which variables and blocks you would like. Unless you'd like to specifically exclude some variables, it is safe just to select all of them (in Paraview you can select or unselect all checkboxes in a group by clicking the checkbox next to the group name). Blocks should all be selected by default, but you can choose not to load some blocks if they are not of interest.
Paraview can optionally be started on the command line, which has the advantage that the file browser will start from your current working directory, which is typically where your output files of interest are located. Search for the path to the executable on your file system, for example, /path/to/paraview. Then rather than specify this full path every time, you can create an alias: alias
paraview='/Applications/paraview.app/Contents/MacOS/paraview'. You could then put this in your ~/.bashrc or equivalent file to have the alias created every terminal session.
Anytime a change is made to the pipeline (opening a file, adding a filter, changing filter properties, etc.), those changes must be applied by clicking the "Apply" button in the properties pane. If you want changes to automatically be applied, you can click the checkbox in Paraview->Preferences.
Viewing 1-D Data
To view 1-D data, do the following:
(Optional) Before viewing any 1-D data, it can be helpful to unload any blocks corresponding to multi-D solutions (click the file name in the pipeline browser and deselect the blocks in the properties pane).
Create a "Plot Over Line" filter:
Filters->Data Analysis->Plot Over Line.Select the begin and end points for the line: in the Properties pane, you'll see "Point1" and "Point2" which are points in 3-D space that define the line to be sampled. As a note, this could also be used to plot a line through a multi-D mesh. If you have only your 1-D meshes, and they are collinear, then the default points will be the ones you want. In general, it selects points to get the smallest bounding box. This is the reason for the optional first step.
Change the sampling resolution: in the Properties pane, there is a box "Resolution", which is the number of points Paraview will take along the defined line. It is recommended to use something quite high, say 1000; otherwise, you may see plotting artifacts. Note that with a high resolution, you will see that the solution is piecewise constant. You could select a resolution that more-or-less matches your number of elements, but you'll probably see a kink in the solution somewhere and falsely think there is a kink in the solution, while in fact it just happens to be where two sampling points finally hit the same element.
Click "Apply".
Close the "Render View" if you do not want it.
Select the variables to view: with the "Plot Over Line" filter in the pipeline selected, select in the properties pane the variables you wish to view.
Created additional plots by clicking the "Split Horizontal" or "Split Vertical" buttons in the upper right corner of a line chart view. Then with a view selected, make sure to activate the eyeball icon next to your "Plot Over Line" filter in the pipeline browser.
Use the green arrows in the header rows to play the transient or step through one time step at a time.
Viewing Multi-D Data
To view multi-D data, do the following:
(Optional) Before viewing any multi-D data, it can be helpful to unload any blocks that do not have the variables of interest. For example, blocks corresponding to 1-D channels will not have the variable
T_solidlike the blocks corresponding to 2-D heat structures. This prevents the scale on color plots from being auto-scaled incorrectly, since a non-existent variable on a block will have the value 0 in Paraview.Select the variable of interest from the dropdown in the header rows (which typically has "vtkBlockColors" selected by default).
Zoom level can be controlled with the scroll wheel.
Use the green arrows in the header rows to play the transient or step through one time step at a time.
Rescale if you wish by using buttons in the header row; there are options for scaling to custom values or to visible values for the current time step.
Some blocks have very small/large aspect ratios, so it can be useful to stretch the coordinate system. After loading a file, with the file selected in the pipeline browser, go to the "Transformation" section near the bottom of the properties pane and use the "Scale" boxes for the x, y, and z directions as needed.
Saving a Paraview State
A Paraview state can be saved with File->Save State. Then you can open up Paraview and automatically load all of the exact same operations you did before with File->Load State (and selecting the .pvsm file you created). There are a number of instances where saving a Paraview state can be useful, for example:
when running an input file multiple times, and
when viewing output while an input file is still running and want to see an update (in this case, you'll want to make sure to reset Paraview first before loading the state by doing
Edit->Reset Session).
If you are opening up Paraview from the command line, you can specify the state file as an argument: