Creator |
|
Language |
|
Publisher |
|
|
Date |
|
Source Title |
|
Publication Type |
|
Access Rights |
|
Related DOI |
|
Related DOI |
|
|
Related URI |
|
|
Relation |
|
|
|
Abstract |
The portmanteau statistic based on the first m residual autocorrelations is used for diagnostic checks on the adequacy of fitting a model with varying m. Katayama (2008) proposed an approximation of t...he joint probability of multiple portmanteau tests with different degrees of freedom (DF). This distribution is easy to compute when all DF are even integers; its empirical behavior is clarified in terms of asymptotic theory. This paper presents detailed simulation study over a variety of ARMA(p, q) models and different m's to show the empirical sizes and powers of the proposed test.show more
|