Hello Asifalig71,

This is an excellent question. We have a few resources available that you may find useful.

For loading data in a matrix I suggest using our new data management tools (introduced in GAUSS 21). We have some recent blogs demonstrating these:

• Easy and Fast Data Management in GAUSS 21
• Preparing and Cleaning FRED Data in GAUSS 21
• Quick and Easy Interactive Data Cleaning Example [GAUSS]

For running GAUSS programs I suggest the following video tutorials:

Dear Erica,
Thank you so much for the information
Best Regards
Benny

Hello,

This type of error occurs because GAUSS is not able to find where the GAUSS procedure PDLMlevel is defined. This likely occurs because the tspdlib library has not been installed properly.

My first suggestion it to reinstall the tspdlib library using the GAUSS package manager.

Erica

Hello Benny,

This test does not account for cross-sectional dependence. However, the PD_nkarul procedure, which applies the panel stationarity test with gradual structural shifts does allow for cross-sectional dependence.

It can also be found in the tspdlib library.

Best,
Erica

I run the data and found this error-G0025 : Undefined symbol: 'PDLMlevel' [code.gss, line 31].

I am new to GAUSS, can someone help please?

Hello Luke,

This is an excellent question and points to some of the issues with panel unit root testing. The situation can arise where, as a panel, we are able to reject the null hypothesis of a unit root but we cannot reject the null for certain individuals within the panel.

The short answer is that:
1. The individual unit root test results tell us what we can conclude for each individual if we consider that individual's behavior alone.
2. The full panel data results tell us what we can conclude for the panel collectively.

This would make an excellent future blog topic!

Erica