about: local variable 'step_size_min' referenced before assignment in pysurvival

[dadekandrew2010]

The error message is prompted by the step function in the rprop.py file in the user torch directory. You only need to initialize the variable that reports the error in this function, similar to step_size_min=[]
def step(self, closure=None):
......
F.rprop(params,
grads,
prevs,
step_sizes,
step_size_min=step_size_min,
step_size_max=step_size_max,
etaminus=etaminus,
etaplus=etaplus)

[mcombalia]

I also have this issue.

[Jaqen00]

I also have this issue.

UnboundLocalError                         Traceback (most recent call last)
Input In [3], in <cell line: 11>()
      8 neural_mtlr = NeuralMultiTaskModel(bins=100, structure=structure)
     10 # Fitting the model
---> 11 neural_mtlr.fit(X_train, T_train, E_train,
     12                 init_method = 'orthogonal', optimizer ='rprop', lr = 1e-4,
     13                 l2_reg = 1e-1,  l2_smooth = 1e-1,
     14                 batch_normalization = True,  bn_and_dropout = True,
     15                 dropout=0.5,  num_epochs = 500)

File ~/PycharmProjects/pysurvival/venv/lib/python3.8/site-packages/pysurvival/models/multi_task.py:369, in BaseMultiTaskModel.fit(self, X, T, E, init_method, optimizer, lr, num_epochs, dropout, l2_reg, l2_smooth, batch_normalization, bn_and_dropout, verbose, extra_pct_time, is_min_time_zero)
    366 Triangle = torch.FloatTensor(Triangle)
    368 # Performing order 1 optimization
--> 369 model, loss_values = opt.optimize(self.loss_function, model, optimizer, 
    370     lr, num_epochs, verbose,  X_cens=X_cens, X_uncens=X_uncens, 
    371     Y_cens=Y_cens, Y_uncens=Y_uncens, Triangle=Triangle, 
    372     l2_reg=l2_reg, l2_smooth=l2_smooth)
    374 # Saving attributes
    375 self.model = model.eval()

File ~/PycharmProjects/pysurvival/venv/lib/python3.8/site-packages/pysurvival/utils/optimization.py:189, in optimize(loss_function, model, optimizer_str, lr, nb_epochs, verbose, num_workers, **kargs)
    187     optimizer.step(closure)
    188 else:
--> 189     optimizer.step()
    190 loss = closure()
    191 loss_value = loss.item()

File ~/PycharmProjects/pysurvival/venv/lib/python3.8/site-packages/torch/optim/optimizer.py:88, in Optimizer._hook_for_profile.<locals>.profile_hook_step.<locals>.wrapper(*args, **kwargs)
     86 profile_name = "Optimizer.step#{}.step".format(obj.__class__.__name__)
     87 with torch.autograd.profiler.record_function(profile_name):
---> 88     return func(*args, **kwargs)

File ~/PycharmProjects/pysurvival/venv/lib/python3.8/site-packages/torch/autograd/grad_mode.py:28, in _DecoratorContextManager.__call__.<locals>.decorate_context(*args, **kwargs)
     25 @functools.wraps(func)
     26 def decorate_context(*args, **kwargs):
     27     with self.__class__():
---> 28         return func(*args, **kwargs)

File ~/PycharmProjects/pysurvival/venv/lib/python3.8/site-packages/torch/optim/rprop.py:109, in Rprop.step(self, closure)
    101         step_size_min, step_size_max = group['step_sizes']
    103         state['step'] += 1
    105     F.rprop(params,
    106             grads,
    107             prevs,
    108             step_sizes,
--> 109             step_size_min=step_size_min,
    110             step_size_max=step_size_max,
    111             etaminus=etaminus,
    112             etaplus=etaplus)
    114 return loss

UnboundLocalError: local variable 'step_size_min' referenced before assignment

[onacrame]

I think this project died so I wouldn't expect an update. Shame as it seemed like a useful package