Plotting with Visclaw¶
- Plotting options in Python
- Using setplot.py to specify the desired plots
- Plotting examples
- Plotting hints and FAQ
- What’s the difference between make .plots and make plots ?
- How to plot a something other than a component of q?
- How to add another curve to a plot, e.g. the true solution?
- How to change the title in a plot?
- How to specify
outdir, the directory to find
fort.*files for plotting?
- How to specify a different
outdirfor some plot item?
- How to set plot parameters that are not provided as attributes of ClawPlotItem?
- How to change the size or background color of a figure?
- Specifying colormaps for pcolor or contourf plots
- How to debug setplot.py?
- GeoClaw plotting tools
- Plotting using Matlab
- Plotting with VisIt
Plotting as post-processing¶
Running a Fortran version of Clawpack produces output files that can then be read in to a graphics package as a post-processing step. If you understand the format of the output files, you can write custom graphics routines using your favorite visualization tools. The output formats are described in the section Output data sytles and formats.
Clawpack includes utilities for plotting using Python, and most of the documention assumes you will use these tools. See Plotting options in Python for a description of these. The Python package matplotlib is used under the hood for 1d and 2d plots, but the tools provided with Clawpack simplify some common tasks since they handle looping over all grid patches as is generally required when plotting AMR data.
Matlab plotting tools (mostly the same as in Clawpack 4.x) are available in Visclaw. See Plotting using Matlab for pointers if you wish to use these tools. For 3d plots the Matlab tools may still be the best choice.
Since Clawpack 4.4, a set of Python plotting tools for 1d and 2d are the recommended approach. The advantages of using the Python options are:
Python and the graphics modules used in Clawpack are open source. Since Clawpack itself is open source we find it desirable to also have an open source plotting open for viewing the results.
The Python tools developed so far (mostly for 1d and 2d data sets) are more powerful than the Matlab versions they replace, and can be used for example to automatically generate html versions of multiple plots each frame over all frames of a computation, to generate latex versions of the output, as well as to step through the frames one by one as with the Matlab tools. It is easier to specify a set of multiple plots to be produced for each frame.
Matlab graphics are somewhat limited for 3d data sets, whereas several open source visualization tools such as VisIt are much better for dealing with large data sets, AMR meshes, etc. and have Python bindings that should allow scripting in a manner compatible with 1d and 2d. (Yet to be done.)
Python is a powerful language that can be scripted to perform multiple runs, such as in a convergence test, and collect the results in tables or plots. We are developing tools to simplify this process.
Plotting on the fly¶