NAG Toolbox for MATLAB®

MBW6I25DDL

Microsoft Windows x64, MATLAB R2015a - R2016a

Supplementary Information



Contents

1. Introduction

This page is for supplementary information relating to the implementation described in Section 2.1 of the Installer's Note. It may be augmented over the life-time of the product if further pertinent information becomes available.

2. R2016b Suitability

This implementation has been tested with MATLAB version 9.1 (R2016b) and is considered suitable for use with that version.

3. R2017a, R2017b, R2018a and R2018b Suitability

This implementation has been tested with MATLAB version 9.2 (R2017a), version 9.3 (R2017b), version 9.4 (R2018a) and version 9.5 (R2018b) and is considered suitable for use with those versions, with the exception of the following routines:

  d03ra / nag_pde_2d_gen_order2_rectangle
  d03rb / nag_pde_2d_gen_order2_rectilinear
  d03rz / nag_pde_2d_gen_order2_rectilinear_extractgrid

4. R2019a, R2019b, R2020a and R2020b Suitability

This implementation has been tested with MATLAB version 9.6 (R2019a), version 9.7 (R2019b), version 9.8 (R2020a) and version 9.9 (R2020b) and is considered suitable for use with those versions, with the exception of the following routines:

  c06pp / nag_sum_fft_realherm_1d_multi_row
  c06pq / nag_sum_fft_realherm_1d_multi_col
  d03ra / nag_pde_2d_gen_order2_rectangle
  d03rb / nag_pde_2d_gen_order2_rectilinear
  d03rz / nag_pde_2d_gen_order2_rectilinear_extractgrid

In our tests, the c06pp and c06pq examples appeared to work on R2020a and R2020b.

In R2019b and later, the example program g13ej_example.m fails due to changes in MATLAB. The underlying routine (g13ej / nag_tsa_kalman_unscented_state_revcom) works as expected once the example is fixed in R2019b and R2020a, but not at R2020b.

On some systems, the Portfolio Optimization demo (Optimization_demo) and the Travelling Salesman demo (travelling_salesman_demo) do not work.

In R2020b, some routines fail with the message:

OMP: Info #273: omp_get_nested routine deprecated, please use omp_get_max_active_levels instead.
Setting the environment variable KMP_WARNINGS=0 works around this problem.