Collection of Python scripts that use "The yt Project" for data manipulation and visualization

Installation can be handled in one of two ways. The new, recommended method which will allow for more native command line usage is to install the entire repo as a python package using pip. This is achieved using the following steps:

conda create -n ytscripts
conda activate ytscripts
conda install python=3.11

The pip package can then be installed either:

git clone https://github.com/nickwimer/ytScripts.git
cd ytScripts
pip install --upgrade .

or directly through github:

pip install --upgrade git+https://github.com/nickwimer/ytScripts.git

NOTE: The pip install will try to install mpi4py which will result in an error if there are no MPI libraries installed on your system. First make sure that the MPI libraries are properly installed.

The older method of installing is to use the environment.yml file to create the conda environment:

To install dependencies, with a working conda distribution, type:

conda env create --file environment.yml

to update an existing conda environment:

conda env update --name ytscripts --file environment.yml --prune

after environment is created:

conda activate ytscripts

NOTE: Most scripts can now make full use of parallel processing either over multiple datasets in a time series or through domain decomposition (depending on the application). Just submit using mpirun -np X (or system equivalent). This is particularly useful when dealing with a large number of time outputs or with very large data.

Documentation is provided through Sphinx under Docs. Everything necessary to build the documentation is provided in the environment.yml file.

Simply navigate to Docs/: cd Docs/

and do make html to build the docs.

Then open Docs/build/html/index.html

Following is a summary of the scripts provided, but this information will be removed in a future update in support for the github docs...

All formatting must pass through black, isort, and flake8

The proper order should be black, then isort, finally flake8 to catch any issues. If there are further changes, repeat the formatting sequence

Scripts for data extraction and manipulation are located under data_extraction/

Extracts 2D slices along specified normal and saves to a npz for later analysis.

Ex: python data_extraction/extract_slices.py -p DATA_DIR/ --field "Y(NC12H26)" --normal x --min 0.005 --max 0.1 --num_slices 16

This will extract all plot files contained in directory DATA_DIR/ for the field Y(NC12H26), making 16 slices starting from x = 0.005 to x = 0.1

These slices will be saved in outdata/slices/ with the following variables: fcoords, normal, iloc, fields, slices, and ds_attributes which contains all attributes specified in utilities.py/get_attributes().

python data_extraction/extract_slices.py --help for full list of arguments.

Extracts domain averaged quantities and saves in a pickled Pandas DataFrame.

The default behavior is to perform a full domain average of the quantity, but 2D averages can be extracted over slices specified with a normal direction and corresponding location keyword.

NOTE: If you have EB boundaries in the domain, you should run with flag --rm_eb to remove the non-fluid regions from the averages.

Ex: python data_extraction/extract_averages.py -p DIR/ --pname plt00001 plt00002 --name NAME --fields mag_vort

Data will be saved under outdata/averages, unless otherwise specified.

python data_extraction/extract_averages.py --help for full list of arguments.

Extracts an isosurface of specified field and value and saves to file for visualization in external program (such as ParaView or Blender).

Ex: python data_extraction/extract_isosurfaces.py -p DATADIR/ --pname plt00001 --field magvort --value 50000.0 --format xdmf

Output formats are ply, obj, hdf5/xdmf.

Isosurface file will be saved under outdata/isosurfaces.

Can be run in parallel with mpi4py. Should run with --do_ghost if there are holes in the iso-surfaces.

Can use --yt to compare the built in iso-surface extraction with the custom version. yt version is not parallelized.

python data_extraction/extract_isosurfaces.py --help for full list of arguments.

Extracts grid information at each level and saves to file.

Ex: python data_extraction/extract_grid_info.py -p DATADIR/ -o OUTDIR/ --name FILE_NAME

Will save a pickle file of the Pandas DataFrame with grid information as a function of time and some stored metadata about the simulation.

Scripts for plotting data that was previously extracted using files under data_extraction/

This will load in the NumPy data structures contained in the input path and make plots.

Ex: python plot_data/plot_slices.py -p outdata/ --field magvort

where outdata is the directory containing the list of .npz files. These images will automatically be saved in imgpath.

python plot_data/plot_slices.py --help for full list of arguments.

• Add in various EB visualization options in quick vis slice plot

This will load in the Pandas DataFrame and make time series plots

Ex: python plot_data/plot_averages.py -p outdata/averages/ -f NAME --field mag_vort

Plots will be saved under outdata/images/

python plot_data/plot_averages.py --help for full list of arguments.

Load grid info data from pickled Pandas DataFrame and make simple plot as an example.

python plot_data/plot_grid_info.py -p DATADIR/ -o OUTDIR/ -f FILENAME

Scripts for in-place quick visualization are located under quick_vis

These scripts are currently provided as examples and will need manual modification. Future updates will make them more automated.

This will take all plt files in the input directory and create 2D images subject to inputs

Ex: python quick_vis/slice_plot.py -p DATA_DIR/ --field "Y(NC12H26)" --normal x --fbounds 0 0.1

This will create a 2D slice plot with x as the normal direction and bounds on field set to 0 - 0.1 in field units

python quick_vis/slice_plot.py --help for full list of options.

Can now make full use of parallel processing over multiple datasets in a time series. Just submit using mpirun -np X or equivalent and images will be processed in an embarrassingly parallel manner.

This script now automatically references a list of configuration settings located in quick_vis/config.toml. If you would like to change any of these default settings simply create a new configuration file called quick_vis/config_user.toml and manually override just the settings that you want to change. The code will update the configuration settings with your new values. NOTE: In the future many of the command line options will be incorporated into these files.

Some helpful options:

--datapath: Path to the plot files.

--outpath: Path to the output image directory (default to datapath/images).

--pname: Used to specify individual plot file names instead of doing all files in the datapath.

--field: Name of the field for visualization.

--normal: Normal direction for the slice plot. Slice will be taken in the middle of the domain.

-c, --center: list of floats for center coordinates if different from center specified with normal.

--fbounds: Bounds for the colormap for the field.

--cmap: Name of the colormap for the slice plot (defualts to "dusk").

--SI: Flag to specify that the data is in SI units (defaults to cgs).

--plot_log: Flag to plot the data using log values.

--grids: Flag to turn on the grid annotations.

--grid_offset: Float value to offset the center value to avoid grid alignment visualization issues (defaults to no offset).

--buff: Buffer resolution for the sliceplot image for plotting.

--dpi: dpi of the output image (default = 300).

--pbox: Bounding box for the output image specified by the two corners of a rectangle (x0 y0 x1 y1).

--contour: Plot a contour line (color=COLOR) of FIELD with VALUE on top of 2D slice. Specified like: --contour FIELD VALUE COLOR.

--clw: Sets the linewidth of the contour lines. Must be same length as the number of contour lines specified with --contour.

--pickle: Optional flag to dump the image as a pickle file for later manipulation/customization.

--grid_info: Add a text box to the slice plot with information about the grids in the simulation. Inputs are: xloc, yloc, min_level, and max_level, where the first two inputs define the location of the text box and the second two inputs define the level range for information in the text box.

--rm_eb: Optional flag to remove the EB boundary from the plot as defined by vfrac field in the dataset. Takes a float to specify the color between [0=white, 1=black].

--gradient: Choice of ["x", "y", "z", "magnitude"] to compute and visualize the gradient of the input field

--no_time: Flag to remove the timestamp in the figure

--no_units: Flag to remove all units from the plots (axes and colorbar)