Compute logistic regression's validation score on the first fold, that is$,$ train on $80\%$ and validate on $20\%$ of the training data $($X$\_$train$)$ without using sklearn Pipeline or cross$\_$validate or cross$\_$val$\_$score. Store the score of the fold in a variable called fold$\_$score. Recall that cross$\_$validation in sklearn does not shuffle the data by default.