EXPERIMENTAL: channel lines based on linear regression and exponential curves (only 3 steps tho) that can be customized to adjust the advance/decay start/middle/end of line plus the lenght of the steps, 2nd step is 4x longer then the 1st.
study(title="[RS]Vector Flow Channel V0", shorttitle="[RS]VFC.V0", overlay=true) decay = atr(10)*0.005 stage1 = input(5) stage2 = input(10) stage3 = input(20) countmax = input(100, type=float) pretopvector01 = nz(topvector01[1], close) topvector01 = close >= pretopvector01 ? close : pretopvector01 - (decay * topvector01finalcounter[1]) topvector01counter = n - valuewhen(close >= topvector01, n, 0) topvector01finalcounter = topvector01counter <= countmax ? stage1 : topvector01counter <= countmax*4 ? stage2 : stage3 prebotvector01 = nz(botvector01[1], close) botvector01 = close <= prebotvector01 ? close : prebotvector01 + (decay * botvector01finalcounter[1]) botvector01counter = n - valuewhen(close <= botvector01, n, 0) botvector01finalcounter = botvector01counter <= countmax ? stage1 : botvector01counter <= countmax*4 ? stage2 : stage3 pt = plot(topvector01, color=(topvector01 == close ? na : black)) pb = plot(botvector01, color=(botvector01 == close ? na : black)) fill(pt, pb, color=blue, transp=90)