search

nbren12 / uwnet

Neural Networks based unified physics parameterization for atmospheric models
MIT License
3 stars 2 forks source link

Data loader for quantiles #52

Open nbren12 opened 5 years ago

nbren12 commented 5 years ago

@sarenehan

  1. Add eta to the xarray training dataset in uwnet.train.get_dataset
  2. subclass: XRTimeSeries. change setup_indices and set time_length=2. ConditionalXrTimeSteps(data: xr.Dataset, conditional_state=2). 
sample = dataset[0]

sample should be

{
'QT': (2, nz, 1, 1)
'stochastic_state': (2, 1, 1, 1)
...
}
nbren12 commented 5 years ago

@sarenehan Since you are no longer training separate NNs for each bin, do you still need these data loader classes? If not, we could close this issue and delete the corresponding branch.