Kaggle & Data Visualisation challenge for Climate Crisis AI 2021 Hackathon developed by the great Vel Ice team
With Random Forests
Overall RMSE: 2.6
With Neural Network
Overall RMSE for velocity magnitude: 4.58
Overall RMSE for velocity direction: 1.28
With XGBoost
Overall RMSE for velocity magnitude: 4.27
Overall RMSE for velocity direction: 1.21
sudo apt-get install libgeos-dev
based on https://stackoverflow.com/questions/42299352/installing-basemap-on-mac-python 1- some home brewing
brew install matplotlib
brew install numpy
brew install geos
brew install proj
2- Download Basemap source tar file (https://sourceforge.net/projects/matplotlib/files/matplotlib-toolkits/), untar it
3- Add "export GEOS_DIR=/usr/local/Cellar/geos/3.5.0/" to a new line in .bash_profile, and then reload it via:
source ~/.bash_profile
4- From within untarred Basemap directory:
python setup.py install
follow https://matplotlib.org/basemap/users/installing.html
Make sure you have python 3 installed as the default version of python then in a command window or terminal run:
python -m pip install -r requirements.txt
run sea_ice_data_plot.ipynb
run sea_ice_data_PCA_analysis.ipynb
run sea_ice_data_visualization.ipynb
To view the plots in the "data visualization" notebook, follow this link:https://nbviewer.jupyter.org/github/Climate-Crisis-AI-Team-Vel-Ice/sea-ice-challenge/blob/master/sea_ice_data_visualization.ipynb
sea_ice_data_ML_XGB_RF.ipynb