Skip to content

Code for future trend prediction of Photosynthetically Active Radiation, written for 2021 NASA Space Apps Challenge: You Are My Sunshine.

Notifications You must be signed in to change notification settings

NASA2021-EcEcursion/POWER_Dips-AzureLSTM

Repository files navigation

POWER_Dips-AzureLSTM

EcEcursion: Andrew Chen, Shu-Yan Cheng, Yun-Hsuan Tsai, Hao Kuan, Yi-Hsuan Lai, Ming-Yi Wei

Code for training in Azure Machine Learning Studio. Trained an LSTM Model with TensorFlow and Keras to predict future trends of Photosynthetically Active Radiation, written for 2021 NASA Space Apps Challenge: You Are My Sunshine.

Packages Used

Keras, Tensorflow, NumPy, Pandas, PlotLy Written with Python 3.6 but may be ok to run with higher versions.

The following should be run in the folder of code to ensure successful build and run:

(Please note that this build is for MacOS/Unix Based Systems)

python3 -m venv AzureLSTM-env
source tutorial-env/bin/activate
sudo pip install tensorflow
sudo pip install numpy
sudo pip install keras
sudo pip install plotly
python Sunshine_LSTM.py

How To Run

Should be able to run directly after ensuring all packages were downloaded in virtual-env of selection Creates .h5 files after training with given files manually downloaded from NASA POWER Project, then moved to AzureServer for further use in the react-native app.

About

Code for future trend prediction of Photosynthetically Active Radiation, written for 2021 NASA Space Apps Challenge: You Are My Sunshine.

Resources

Stars

Watchers

Forks

Packages

No packages published