Files: cnn.py - python script to run neural network README.md - this README file
Example Data files: h1waves_test_csv - h1 component for testing data h1waves_train.csv - h1 component for training data h2waves_test.csv - h2 component for testing data h2waves_train.csv - h2 component for training data parrivals_test.csv - p arrival picks for testing data parrivals_train.csv - p arrival picks for training data sarrivals_test.csv - s arrival picks for testing data sarrivals_train.csv - s arrival picks for training data timing_test.csv - waveform timing for testing data timing_train.csv - waveform timing for training data Zwaves_test.csv - vertical component for testing data Zwaves_train.csv - vertical component for training data
Outputs: figures - figures included in final paper results - outcome figures for running neural network (will overwrite if code is run) predicted_model.npy - python numpy file for the network output arrays. These are probability outcomes for p and s arrivals for each testing waveform. weight.h5 - neural network model weights found from the fitting training data to model
To run cnn.py the following python packages are required in addition to the regular Anaconda 3 installation:
- tqdm
- skimage
- keras
- tensorflow