Let f:R+→R, and π=(t0,…,tN). We define the variation of f on the subdivision π, V(f,π), as V(f,π)=i=0∑N−1∣f(ti+1)−f(ti)∣or the variation of f over [a,b] as V(f∣[a,b])=π∈Π([a,b])supV(f,π)
Let 0≤a<b, and let f:R+→R, f is said to have bounded on [a,b] if π∈Π([a,b])supV(f,π)<∞where Π([a,b])={(t0,…,tN)∈RN+1:N∈N∗ and a=t0<t1<⋯<tN=b}for π=(t0,…,tN).
Let f:R+→R. f is said to have finite variation if f∣[0,t] is of .