You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I use 32 columns of dummy variable to indicate industry, and concat with 10 factors I've got from the model. I use SecuCode and TradingDay as multiindex.
However during the training it keeps erroring like this:
in calc_factor_ret
cap_w = stock_cap.values @ X[:, slc]
~~~~~~~~~~~~~~~~~^~~~~~~~~~~
ValueError: matmul: Input operand 1 has a mismatch in its core dimension 0, with gufunc signature (n?,k),(k,m?)->(n?,m?) (size 3461 is different from 1)
If i fix this by adding a .T to stock_cap, there will still be other errors. I guess it's because the forms of data we use are different. Could you please share me with your data's form?
The text was updated successfully, but these errors were encountered:
Liu-200106
changed the title
torch.inverse Error: Singular Matrix during Training
original forms of data issue
Jun 21, 2024
I use 32 columns of dummy variable to indicate industry, and concat with 10 factors I've got from the model. I use SecuCode and TradingDay as multiindex.
However during the training it keeps erroring like this:
in calc_factor_ret
cap_w = stock_cap.values @ X[:, slc]
~~~~~~~~~~~~~~~~~^~~~~~~~~~~
ValueError: matmul: Input operand 1 has a mismatch in its core dimension 0, with gufunc signature (n?,k),(k,m?)->(n?,m?) (size 3461 is different from 1)
If i fix this by adding a .T to stock_cap, there will still be other errors. I guess it's because the forms of data we use are different. Could you please share me with your data's form?
The text was updated successfully, but these errors were encountered: