This function is used to find the potential trend turning point in long
term datsets.
Code written by Pankaj Dey, PhD scholar, Department of Civil Engineering,
Indian Institute of Science, Bangalore.

Input: The time series data with first column with time index and the
second column with the values of the variable. Size is n x 2, where n is
no. of elements in time series dataset.


Output: 'sqmk' consists of 3 columns. The first column gives the
progressive series, second column gives the retrograde series and the
third one gives the logical series. The ones are the value in the time
index from where the potential truning point with respect ot trend has
occured.


Reference: Chatterjee, S., Bisai, D. and Khan, A., 2014. Detection of approximate potential trend turning points in temperature time series (1941–2010) for Asansol Weather Observation Station, West Bengal, India. India. Atmospheric and Climate Sciences, 4, pp.64-69.