Cookbook

Installation

Install under Python 3. Please note that under Python 3 it is no longer possible to create a 3D view of the flake due to the lack of the mayavi package. Switch to main directory and..

... first install requirements...

$ pip install -r requirements.txt

... and then the flame module:

$ python setup.py install

Single Flake

This is run in a python console or as script.

• Create the Flake: Import the growth module e.g. from flame import growth and create an instance of the Flake: myFlake = growth.Flake()
• Inspect the Flake: To view basic information, the representation of myFlake will display it's properties (twinplanes, seed, iterations, etc.). A visual representation is rendered in mayavi through myFlake.plot()
• Grow atoms: Now we grow 20 atoms by calling the grow method on the Flake and passing the number as argument myFlake.grow(20). The growth method can be refined through various arguments, particularly the mode can be specified. As default we will grow with an exponential probability distribution according to our temperature in myFlake.temp.
• Export results(OUTDATED): After several iterations of inspecting and growing the Flake you can save the picture by passing save=True to the plot() method. This is then exported in output directory of the corresponding date. To save the coordinates in a xyz-file call the export method, with a describing tag: myFlake.export_coordinates('unexpected_phenomenon') The resulting file is plaintext containing the coordinates and can be used to import the coordinates into Blender for rendering or further use.

Automated Flake generation

This is run in a shell calling the program with arguments.

• Create a simulation: - create new directory with the desired name - change into the directory - run $ flame create This will create a new file name sim_params.yaml containing conservative default settings for a new simulation. Those can be edited as needed. The most relevant is the function and the values for creating twin planes. This is used to sweep this parameter as desired.
• Run simulation: After changing the settings to your desire run $ flame run and wait for the simulation to run. A HDF5 file is created with the name of the project and a random salt to make it possible to run multiple simulations with the same parameters. The simulation grows flakes with the given parameters and at each growth step defined by snapshot interval the output of the geometry() method is then saved.
• View results: After the simulation is done, run $ flame paint <column1> <column2> <...> in the project folder. Where the columns are the aspects of interest of the geometry() method. Each result is then rendered and saved to a file named <column>.html and can be viewed, scrolled and panned in a regular browser.