A calculator gives you a quick numerical answer; Julia also lets you inspect what kind of number produced it. Floating-point numbers come in different bit lengths. The default, Float64, uses 64 bits (8 bytes). More bits can support larger values and higher precision, though a graphics calculation may need less precision than a scientific one. section 2.0 section 2.4
julia> 42.98 #1 42.98 julia> typeof(ans) Float64 julia> 34.23f0 #2 34.23f0 julia> typeof(ans) Float32
Compare the types of `42.98` and `34.23f0`: an ordinary decimal literal is Float64, while the `f0` suffix makes a Float32 literal. section 2.4
The zero in `f0` matters. Unlike an `f` suffix in Java or C/C++, `0.5f` in Julia is interpreted as multiplication by a variable named `f`. For a 16-bit value, convert explicitly with `Float16(3.5)`. Types can also change across an operation: `Float16(4) + 5f0` produces a Float32 result. section 2.4
Choose the result you need
In the REPL, `4/2` returns `2.0` and `5/2` returns `2.5`: `/` produces a floating-point result even for integer inputs. If you need the integer quotient and remainder instead, use `div(5, 2)` for `2` and `rem(5, 2)` for `1`. section 2.4.1
Rounding is a separate choice. `floor(3.6)` returns `3.0`, `ceil(3.2)` returns `4.0`, and `round(3.5)` returns `4.0`. These results are still floating-point values; use `round(Int64, 3.5)` when you want an integer result. section 2.5.4
