Saving break years after structural break test

[pras]

Hello everyone,

I am very new with EViews. Can you help me how to save the years from the structural break test? Here is my simple code and attached is the sample data. The output that I aim is dummy variable for the years when structural break take place for each country/variable. Thank you.

Code: Select all

%country = "BGD IND LKA MDV PAK" ' Loop through each dependent variable and run the regression for %y {%country} ' Create an equation object with a dynamic name equation eq_{%y} ' Estimate the regression eq_{%y}.ls {%y} c year ' Run multibreak test eq_{%y}.multibreak c year next

Best regards,

Pras

[EViews Gareth]

You could estimate via breakls instead of using the breaktest, and then store the break points.

Code: Select all

%country = "BGD IND LKA MDV PAK" ' Loop through each dependent variable and run the regression for %y {%country} equation eq_{%y}.breakls {%y} c year string {%y}_breaks = eq_{%y}.@breaks next

[pras]

Hello Gareth,

Thank you your help. Is there any way I can export the string into excel? Ideally I need to have an excel table where the rows are the year and the columns are each country/variable, with value = 1 if the year is equal to the break year. If that is not possible with EViews, at least I need to combine all the resulting string into one group and then export them to excel. At the moment, I am manually adding the break year dummy in excel, but it's time consuming and error-prone. Sorry if my question is too basic. Thank you.

[EViews Gareth]

You can make a table object, insert the information you want into the table object, then save the table to disk as an Excel file, or csv.

[pras]

Thanks, Gareth.

Do I have to fill the table manually as in my simple code below, or is there any way to automate which rows to be changed based on the structural break years? For the way I did it below, it's still time consuming and error-prone. Thanks.

Code: Select all

' Creating the table ' Step 1: Define the number of years scalar startYear = 1990 scalar endYear = 2022 scalar numYears = endYear - startYear + 1 ' Step 2: Create a table with the appropriate number of rows and 6 columns table (numYears + 1, 6) mytable ' +1 for the header row ' Step 3: Name the columns mytable(1, 1) = "index" mytable(1, 2) = "year" mytable(1, 3) = "BGD" mytable(1, 4) = "IND" mytable(1, 5) = "LKA" mytable(1, 6) = "MDV" mytable(1, 7) = "PAK" ' Step 4: Fill the "Year" column with the range of years for !i = 1 to numYears mytable(!i + 1, 1) = !i ' Fill the index column mytable(!i + 1, 2) = startYear + !i - 1 ' Fill the year column next ' Filling the break years dummy for BGD mytable(13, 3) = "1" mytable(20, 3) = "1" mytable(24, 3) = "1" mytable(28, 3) = "1"

[EViews Gareth]

Entirely depends on what you want it to look like in Excel. But the simplest is probably something like:

Code: Select all

table mytable %country = "BGD IND LKA MDV PAK" !row = 2 ' Loop through each dependent variable and run the regression for %y {%country} equation eq_{%y}.breakls {%y} c year %breaks = eq_{%y}.@breaks mytable(!row,1) = %y !col = 2 for %j {%breaks} mytable(!row, !col) = %j !col = !col+1 next !row=!row+1 next

[pras]

Thanks a lot, Gareth.

If there are no break years, the code returns an error. I think I need to modify this code:

Code: Select all

!col = 2 for %j {%breaks} mytable(!row, !col) = %j !col = !col+1 next

I tried the following, but still returning error. Could you please help, Gareth? Thanks a lot!

Code: Select all

!col = 2 if {%breaks} <> "" then for %j {%breaks} mytable(!row, !col) = %j !col = !col+1 next else mytable(!row, !col) = "" endif

[pras]

Hello Gareth,

I think I found the solution as follows. Thanks for your help.

Code: Select all

!col = 2 if @len(%breaks) > 0 then for %j {%breaks} mytable(!row, !col) = %j !col = !col+1 next else mytable(!row, !col) = "" ' Fill empty cells in case of no breaks endif