SpecEval add-in: Add-in for time series model development
Posted: Thu Apr 15, 2021 5:32 am
This is post for SpecEval add-in. The add-in is a multi-task add-in facilitating (time series) model development.
The add-in performs evaluation of (forecasting properties of) equation/VAR/identity object(s) and prepares report summarizing the results. It will return spool with tables and graphs that can be used for evaluation of (forecasting) performance of given equation, and/or compare forecasting performance across alternative specifications.
The add-in is aimed at facilitating iterative and interactive model development process, in which model developer improves the model based on information about the entertained specification(s). Consult separate document ‘SpecEval illustrated.pdf’ for illustration of model building process and for multiple examples of use of the add-in. (Associated data and programs are available on my GitHub site).
The add-in provides great deal of control to the user, who can adjust the procedures and outputs along many different dimensions. All options are documented in the add-in documentation, and most are illustrated in applications in separate document. Similarly, all the outputs and procedures are documented in dedicated document.
This is a live project; if you want to contribute feel free to reach out either here, or directly on my GitHub site.
The add-in performs evaluation of (forecasting properties of) equation/VAR/identity object(s) and prepares report summarizing the results. It will return spool with tables and graphs that can be used for evaluation of (forecasting) performance of given equation, and/or compare forecasting performance across alternative specifications.
The add-in is aimed at facilitating iterative and interactive model development process, in which model developer improves the model based on information about the entertained specification(s). Consult separate document ‘SpecEval illustrated.pdf’ for illustration of model building process and for multiple examples of use of the add-in. (Associated data and programs are available on my GitHub site).
The add-in provides great deal of control to the user, who can adjust the procedures and outputs along many different dimensions. All options are documented in the add-in documentation, and most are illustrated in applications in separate document. Similarly, all the outputs and procedures are documented in dedicated document.
This is a live project; if you want to contribute feel free to reach out either here, or directly on my GitHub site.