search

UPennEoR / ml_rfi

Machine learning applied to RFI flagging
BSD 2-Clause "Simplified" License
8 stars 4 forks source link

Refactor the train, evaluate, test options #6

Closed jkerrigan closed 5 years ago

jkerrigan commented 6 years ago

Train, evaluate, and test are currently jumbled around into a bunch of different if statements. This could definitely be sorted properly as the original reason for doing this was how the simulated and real datasets were partially merged with the remaining split for the evaluation mode. Maybe a function that just takes in the two datasets and splits it accordingly between training and evaluation would be useful.

jkerrigan commented 5 years ago

Handled in the cleanup branch with these options being divvied up appropriately to different scripts: runDFCN.py and pipelineDFCN.py