NEWS.md

In nortsTest: Assessing Normality of Stationary Process

Packages NEWS and Updates

nortsTest 1.1.2 Date: 25/01/2024

Fixes:

• Minor bug fix with seed when running rp tests using parallel programming.
• Minor bug fix with seed when running bootstrap tests using parallel programming.

nortsTest 1.1.1 Date: 23/01/2024

Fixes:

• use cowplot package for the check_residuals plot function.
• Update rejection_rate and rejection_table functions to be more efficient.
• Increase time compilation checks for test-performance to 30s and test-basics to 15s.
• change the rp.sample function. It computes 2k projections and apply epp.statistic to odd projections and lobato.statistics to the even ones.

nortsTest 1.1.0 Date: 16/01/2024

Features:

• Add initial values as an argument to the Epps and Pulley test.
• Add the bivariate elbouch.test().
• Add the lobato-bootstrap.test().
• Add the epps-bootstrap.test().
• Add the jb-bootstrap.test().
• Add the shapiro-bootstrap.test().
• Add the cvm-bootstrap.test().
• Add unit testing for a better debug practices.

Fixes:

• Update documentation for Lobato, RP, Epps, and Vavra tests. Better description to the return value, having a similar formatting to the t.test() function.
• Update documentation for normal.test, seasonal.test, and uroot.test. Better description to the return value, having a similar formatting to the t.test() function.
• Fix the htest print method for the random projections test. The function does not print the average Epps and Lobato's statistics.

Changes:

• Use Hochberg's False discovery rate method as default to mix p.values, when applying the rp.test().
• Refactor the lobato.statistic() using less for loops.
• Refactor the epps.statistic() using less for loops.
• Speed up vavra.test() and rp.test() by replacing for loops with parallel vectorized computation.

nortsTest 1.0.3 Date: 12/06/2021

Features:

Changes:

Fixes:

• Fix documentation.
• Update p-values mixing method using a False discovery rates, and Benjamin and Yekutieli (2001) procedures.
• update the random projections procedure for the rp.test.

nortsTest 1.0.1 Date: 17/06/2021

Features:

• Add GPL-2 license.

Changes:

• False discovery rate set as default for the random projections test

• Change "NA %in$ y" to "anyNA(y)"

• use match.args() function for more flexibility

Fixes:

• Bug fix for the False discovery rate.

nortsTest 1.0.0 Date: 08/07/2020

Fixes:

• Bug fix for CRAN submission.

nortsTest 1.0.0 Date: 12/06/2020

Fixes:

• Bug fix for CRAN submission.

nortsTest 1.0.0 Date: 09/05/2020

Features:

• vavra.test() function for the Psaradakis and Vávra test.

• sieve.bootstrap() function for bootstrap sub-sample in stationary time series.

• vavra.sample() function for the Anderson Darling sample statistics for the Psaradakis and Vávra test.

• rp.test() function for the random projections test.

• rp.sample() function for the random projections statistics samples.

• random.projection() function to generate a random projection of a stationary process.

• gghist(), ggnorm(), ggacf() and ggpacf() function for visualization.

Improvements:

• epps.statistic() bug fixed.

• package documentation, errors, warnings and notes corrected.

Changes:

• plot.compare() function deleted

• epps.statistic() using the PoweR package deleted.

Fixes:

• The amoebam algorithm for epps.statistic() is corrected.

normality 0.0.1.000 Date: 12/03/2020

Features:

• normal.test() function with all the normality test available

• uroot.test() function with all the unit root test available

• seasonal.test() function with all the seasonal unit root test available

• arch.test() function with the arch effect test for stationarity

• check_plot() methods implemented for residual diagnostics visualization for lm, glm, ets, forecast, Arima, arima0, fgarch, HoltWinters, ts, and numeric classes

• check_residuals() methods implemented for residual diagnostics for lm, glm, ets, forecast, Arima, arima0, fgarch, HoltWinters, ts, and numeric classes

• the autoplot() methods are overloaded for plotting time series (ts) and multivariate (mts) classes

Improvements:

• lobato.test() is implemented as a htest method

• epps.test() is implemented as a htest method

Changes:

• LV.statistic() function change to lobato.statistic()

Fixes:

• The amoebam algorithm for Epps.statistic() is changed for the one implemented by the PoweR package

normality 0.0.1.000 Date: 23/10/2020

Features:

• Updated the epps.statistic function

• Updated the LV.statistic function

• Perform documentation and package description

normality 0.0.1.000 Date: 22/10/2020

Features:

• Package Repository created

• Discussion of the previous work and code homogeneity check README.md

• Incorporation of references and latex thesis algorithm

• Incorporation of the article structure

nortsTest documentation built on May 29, 2024, 10:05 a.m.