It's Christmas let's go home! The main goal of our project is to reproduce the results of this paper on snowflake growth simulation in an object-oriented C++ code. We also implemented a box-counting algorithm to calculate the fractal dimension of snowflakes boundaries in Python. CMake and a C++ compiler (e.g. g++) are required to compile and produce snowflakes.
You must first clone the repository recursively to include the JSON parsing submodule RapidJSON:
git clone --recursive
For Unix-like and Mac users, the bash script compile.sh can be run to compile the project and move the binary executable into the root of the project. For Windows, the following commands will do the same
mkdir build
cd build
cmake ..
make
mv ./snowflake ../snowflake
cd ..
Once the code is compiled, you can input the parameters of the snowflake in a JSON configuration file (you can see an example in parameters.json). Simply run:
./snowflake parameters.json
The program saves the snowflake into four binary matrices: vapor mass (double), solid mass (double), liquid mass (double) and the snowflake's interior (boolean). If specified in the configuration file, the attribute frameskips allows for animated gifs: a frame (i.e. the snowflake) is saved at every number of frames skipped. Every frame is named filepath_prefix_<iteration>_<snowflake atttribute> and the final file is named filepath_prefix_<snowflake atttribute>.
The Python file hex_grid.pycontains a function to plot/export the different masses on the domain. A hexagonal grid is used to represent the underlying crystalline structure. A snapshot of the snowflake can also be exported.
Using the script fractal_dimension.py, you can get the box-counting dimension of a curve in 2D space. We recommend using a liquid mass file to compute the fractal dimension of a snowflake, since liquid is located only at the boundary, i.e. the liquid mass curve is a good approximation of the snowflake's boundary. Some of the parameters of the script have to be modified according to the size of the data frame.