Percent Change Smoothed (PCT + EMA)

2018年10月11日

ROC works great on data with only positive numbers (like prices).
But it fails to correctly represent the rate of change when source series have negative values.

ROC is positive when:

• Source is positive AND Source is rising
  
• OR Source is negative AND Source is falling
  

Percent change (PCT) is just a ROC that deals with this sign confusion.
Anyone with Data Science backgroud would likely know about it.

PCT is positive only when:

• Source is rising
  

When applying to only positive data PCT = ROC they are exactly equal.

I've also added EMA smoothing option.

Enjoy!