im trying to use correlation matricies in r my data has bunch of null or n/a values my current approach is to convert these null values to 0 this works but it results in an inaccurate matrix because some columns that have correlation are overpowered by the 0 do you know of any solution to fix the na values?