I have an equation that involves lots of divisions to produce a number (say) "X" then raising "X" to the power 0.5 (square rooted) gives me the final answer "Y". My question is: If I use BigDecimal to do the divisions and multiplications until I get "X" then I convert "X" to double in-order to use the .pow() method to get Y, (I know I might loose precision in the last step) but how bad is that idea? I am saying so because I have seen people having troubles in raising a BigDecimal to a double value.
Any recommendations/corrections are welcome.