#Divide each column by its number of stories to get the mean
hourly_means = pd.DataFrame({col: hourly_mi[col] / hourly_mi['count'] for col in hourly_mi.columns}).fillna(0)