2. Use sklearn.preprocessing package to process data. Particularly, I use the function called ‘MinMaxScaler’ to ‘scale’ leading cause of death data (count of each death, after I remove the top three and combine the metal health factors). It produces a data matrix only between 0 and 1, yet I got some ‘nan’s in the very last couple of rows and I could not figure out why (is it because the difference between max and min after processed is still too big?) Comparing to simply writing a function that would also normalize the data, I find this method is little confusing but it can be used to perform really neat figure.