PGPortfolio: Policy Gradient Portfolio, the source code of "A Deep Reinforcement Learning Framework for the Financial Portfolio Management Problem"(https://arxiv.org/pdf/1706.10059.pdf).
In training phase, I get a assertionerror:
File "E:\github\PGPortfolio\pgportfolio\learn\nnagent.py", line 153, in train
self.evaluate_tensors(x, y, last_w, setw, [self.__train_operation])
File "E:\github\PGPortfolio\pgportfolio\learn\nnagent.py", line 169, in evaluate_tensors
"the last_w is {}".format(last_w)
AssertionError: the last_w is [[0.09090909 0.09090909 0.09090909 ... 0.09090909 0.09090909 0.09090909]
[0.09090909 0.09090909 0.09090909 ... 0.09090909 0.09090909 0.09090909]
[0.09090909 0.09090909 0.09090909 ... 0.09090909 0.09090909 0.09090909]
...
[ nan nan nan ... nan nan nan]
[ nan nan nan ... nan nan nan]
[ nan nan nan ... nan nan nan]]
Has anyone encountered this problem and how is it solved？
In training phase, I get a assertionerror: File "E:\github\PGPortfolio\pgportfolio\learn\nnagent.py", line 153, in train self.evaluate_tensors(x, y, last_w, setw, [self.__train_operation]) File "E:\github\PGPortfolio\pgportfolio\learn\nnagent.py", line 169, in evaluate_tensors "the last_w is {}".format(last_w) AssertionError: the last_w is [[0.09090909 0.09090909 0.09090909 ... 0.09090909 0.09090909 0.09090909] [0.09090909 0.09090909 0.09090909 ... 0.09090909 0.09090909 0.09090909] [0.09090909 0.09090909 0.09090909 ... 0.09090909 0.09090909 0.09090909] ... [ nan nan nan ... nan nan nan] [ nan nan nan ... nan nan nan] [ nan nan nan ... nan nan nan]] Has anyone encountered this problem and how is it solved？