check from the website https://github.com/rvlasveld/ICSS/blob/master/README.md

it can implemented using matlab.  Hope it helps you

Thank you very much for the link I have already done the calculations in an Excel file based on Inclan C.and Tiao G.C.,1994*. I will use the provided link to make sure I got it right

thank you very much

* Inclan C.and Tiao G.C.,1994, Use of Cumulative Sums of Squares for Retrospective Detection of Changes of Variance, Journal of the American Statistical Association, Vol. 89, No. 427 (Sep., 1994), pp. 913-923

I still have one more question, I get several break points in different dates  and I will create a  dummy variable to represent each one. In literature it is stated that the dummy should be equal to (1) at the break point  onward and (0) otherwise. Should the dummy variable continue in the value (1) till the end of the data or just till the beginning of a new  break point ?

for example, lets use the economic liberalisation in ghana which started in 1983. for you to use a dummy to represent this structural break in a data set from let say 1960 to 2013. So you set the period after 1984 (ie 1984 to 2013) as (1) and periods before 1984 (ie 1960 to 19853) as (0). 

the individual breaks you see are caused by a policy change so it affects all the period unless there are different policies that were implemented within the study frame. for you create individual dummies.

example, have 2 structural breaks caused by three exchange rate policy reforms from 1960 to 2013.

first,  from 1960 - 1980 = the economy was using crawling exchange rate

second, from 1981 - 2000= the economy was using flexible

with these two reforms you create the following individual dummies in the form of variables

1960-1980: dummy1  = 1 if reform is between 1960 - 1980, 0 otherwise

1981 - 2000: dummy2 = i if reform is between 1981 - 2000, 0 otherwise  

Thank you very much for the answer. I know this for the exogenously detected breaks like the examples you have kindly wrote. In the case of ICSS structural breaks are endogenously detected which means we can know when the effect of the shock was removed by looking at the plot of the Dk  statistic against time . My calculations are in the attached Excel file , it can be examined together with the cited article (Inclan C.and Tiao G.C.,1994)

I really appreciate the answer and I will follow the same approach since there is no evidence for a different approach  related to endogenously detected structural breaks

Check this paper. he used similar modelling technique as you are using. In it he has introduced some dummies which I believe would help you catch up. it is found just before his references

Thank you very much for the reference

For the application of ICSS algorithm I found out it can be done using WINRATS package which can be downloaded (15 days trial) for free from estima.com

https://estima.com/ratsdemo.shtml

thanks great. wish you all the best

@ Mazen Alaboud.... how you plot graph in excel for Dk and did you created dummies from graph created in excel? for dummies creation which bench mark you used.al these questions are with reference to excel file you uploaded. although the entire work regarding ICSS algorithm you did is in RATS but it will be very helpful if you answer above questions according to that particular work you done in excel.

A quick reply is expected.

Hi Sir Mazen Alaboud,

How did you calculate resid01 (Col C in your Excel file)

I hope to be replied soon

Thank Sir

Linh