The tan command computes the tangent of x where x is in radians. If x is a floating point number then a floating point result is computed, otherwise the result is an exact symbolic value.
# hardware floating point tan(1.0); 1.5574077246549 # symbolic tan(1); tan(1)