Regarding the Logisitc Regression model
When attempting to fit the Logistic Regression model from the previous video, there is a type error indicating that there are mixes of strings and integers even when using direct downloads from the resources. In addition, even after fixing the datatypes of the loan_data_targets_train variables so they are all integers, when trying to run the fit method again, it somehow reaches a max amount of iterations unlike within the video. Is something wrong here or do I need to include the max_iters argument?
4 answers ( 0 marked as helpful)
Yeah I am also facing the same problem.
checkk the missing values
I have the same issue. Has anyone been able to resolve it, please?
1) In the fit method, update the line as
"F_ij = np.dot((X / denom).T,X).astype('float')"
2) and when invoking the class instance pass the parameter as "reg = LogisticRegression_with_p_values(solver='liblinear')".
3) update the line as "reg.fit(inputs_train, loan_data_targets_train.values.ravel())".
This worked for me.
"F_ij = np.dot((X / denom).T,X).astype('float')"
2) and when invoking the class instance pass the parameter as "reg = LogisticRegression_with_p_values(solver='liblinear')".
3) update the line as "reg.fit(inputs_train, loan_data_targets_train.values.ravel())".
This worked for me.