write a python code

you have a dataset of food recipes labeled with their primary nutritional category $($e$.$g$.,$ "high$-$protein," "low$-$carb," "balanced," etc.$).$ Each recipe is represented by a feature vector indicating the relative proportions of its key ingredients. Train a KNN model using ingredient proportions, setting k to the square root of $4.$ Use $10-$fold cross$-$validation and report the accuracy, precision, recall, F$-$measure, and AUC for each nutritional category.

Part b: Explore automating the process of finding the optimal k value for your KNN model by modifying its parameters