GitHub - jkerrigan/RFINeural

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 19 Commits
zen.2457691.54970.xx.HH.uvcA		zen.2457691.54970.xx.HH.uvcA
CNNRFI.py		CNNRFI.py
CNNvsRFI.ipynb		CNNvsRFI.ipynb
DEEPcnn.txt		DEEPcnn.txt
FindHypers.py		FindHypers.py
NeuralAssignPredict.py		NeuralAssignPredict.py
PaintRFINeuralTrain.py		PaintRFINeuralTrain.py
README.txt		README.txt
apply_xrfi.py		apply_xrfi.py
basicNN.py		basicNN.py
cnn.txt		cnn.txt
heranet.txt		heranet.txt
trainMask_HQ.txt		trainMask_HQ.txt
trainNN.py		trainNN.py
xrfi.py		xrfi.py

Repository files navigation

CNNRFI.py - Main script used to train the CNN, requires access to a larger dataset for incorporating more baseline types.

FindHypers.py - Extension of CNNRFI.py that uses the same CNN structure to iterate over the parameter space and find the optimal parameters.

NeuralAssignPredict.py - Takes the output model (DEEPcnn.txt) and applies it to the rest of the dataset. Usage is 'python NeuralAssignPredict.py zen.2457691.54970.xx.HH.uvc', output file will be 'zen.2457691.54970.xx.HH.uvcr'. To view output information however requires the installation of AIPY at https://github.com/AaronParsons/aipy. To view output RFI flagged file then run 'plot_uv.py zen.2457691.54970.xx.HH.uvcr -a 9_31', which will show the log amplitude of baseline with antennas 9 and 31.