Test Full Run Speed (Savings Distribution)

back to Fan's Dynamic Assets Repository Table of Content.

Testing the ff_az_ds_vecsv program for solving the savings only dynamic programming problem.

Here the dynamic programming problem is solved and stationary distribution is found. This when invoked multiple times, lead to finding the equilibrium prices. See Wang (2019)'s Appendix for how to solve equilibrium prices in this context using multile cores. Specifically, if there is only one processor, the code could be invoked up to 12 to 15 times to find the interest rate using bisection. If the computer has several processors, equilibrium could be found using multi-section as described in Wang (2019) in 2 or 3 iterations where the code is run concurrently across multiple processors within each iteration.

@seealso

SPEED savings only overall benchmark speed testing: fsi_az_ds_vecsv_speed
PREFERENCE savings only preference testing: fsi_az_ds_vecsv_pref
PREFERENCE savings only preference testing CROSS: fsi_az_ds_vecsv_pref_cross
PREFERENCE savings only preference testing GRID: fsi_az_ds_vecsv_pref_grid
SHOCK savings only shock testing: fsi_az_ds_vecsv_shock
SHOCK savings only shock testing CROSS: fsi_az_ds_vecsv_shock_cross
SHOCK savings only shock testing GRID: fsi_az_ds_vecsv_shock_grid
PRICE savings only wage and interest rate testing CROSS: adjust wage and savings rate fsi_az_ds_vecsv_price_cross
JOINT all parameters random draws joint test fsi_az_ds_vecsv_joint_rand

Solving the Benchmark Model

Solving the Benchmark Model

close all;
clear all;

% Start Profiling
bl_profile = false;
if (bl_profile)
    profile off;
    profile on;
end

% Start Timer
bl_time = true;
if (bl_time)
    tic;
end

% Set Parameters
bl_input_override = true;
it_param_set = 9;
[param_map, support_map] = ffs_az_set_default_param(it_param_set);
support_map('bl_time') = false;
support_map('bl_display_final_dist') = false;

% Call Grid Generator <https://fanwangecon.github.io/CodeDynaAsset/m_az/paramfunc/html/ffs_az_get_funcgrid.html ffs_az_get_funcgrid>
[armt_map, func_map] = ffs_az_get_funcgrid(param_map, support_map, bl_input_override);

% Call Dynamic Programming Problem <https://fanwangecon.github.io/CodeDynaAsset/m_az/solve/html/ff_az_vf_vecsv.html ff_az_vf_vecsv>
result_map = ff_az_vf_vecsv(param_map, support_map, armt_map, func_map);

% Call Distribution CProgram
result_map = ff_az_ds_vecsv(param_map, support_map, armt_map, func_map, result_map, bl_input_override);

% End Timer
if (bl_time)
    toc;
end

% End Profiling
if (bl_profile)
    profile off
    profile viewer

    % append function name
    st_func_name = 'fsi_az_ds_vecsv_speed';
    support_map('st_profile_path') = [support_map('st_matimg_path_root') '/test/ff_az_ds_vecsv/test_speed/profile/'];
    support_map('st_profile_name_main') = [st_func_name support_map('st_profile_name_main')];

    % support_map
    params_group = values(support_map, {'st_profile_path', ...
        'st_profile_prefix', 'st_profile_name_main', 'st_profile_suffix'});
    [st_profile_path, st_profile_prefix, st_profile_name_main, st_profile_suffix] = params_group{:};

    % Save
    st_file_name = [st_profile_prefix st_profile_name_main st_profile_suffix];
    profsave(profile('info'), strcat(st_profile_path, st_file_name));
end

Elapsed time is 5.510150 seconds.

Test Full Run Speed (Savings Distribution)

Contents

Solving the Benchmark Model