Trusted, Bureaucrats
1,382
edits
Changes
* To return to follow mode, use keyboard: '''Shift+f'''
* To exit, use keyboard: '''Ctrl+c''' and then: '''q'''
=== Try normal run with MC 20 and MAX_ITER 20 ===
It looks like the '''prolate''' has problem with converging. <br>
So let us try a run, where a maximum of 20 rounds of convergence is accepted
We have 4 CPU on our lab computers.<br>
Let us assign another to a run normal settings, only MC=20 and MAX_ITER=20.
Copy '''05_run_def_MC20.py''' to '''06_run_def_MC20_MAX_ITER20.py'''
<source lang="bash">
cp 05_run_def_MC20.py 06_run_def_MC20_MAX_ITER20.py
</source>
and modify last lines:
<source lang="python">
# The number of Monte Carlo simulations to be used for error analysis at the end of the analysis.
#MC_NUM = 500
MC_NUM = 20
# The diffusion model. Standard is 'Fully automated', which means: DIFF_MODEL=['local_tm', 'sphere', 'prolate', 'oblate', 'ellipsoid', 'final']
# 'local_tm', 'sphere', ''prolate', 'oblate', 'ellipsoid', or 'final'
#DIFF_MODEL = 'local_tm'
DIFF_MODEL = ['local_tm', 'sphere', 'prolate', 'oblate', 'ellipsoid', 'final']
# The maximum number of iterations for the global iteration. Set to None, then the algorithm iterates until convergence.
MAX_ITER = 20
# Automatic looping over all rounds until convergence (must be a boolean value of True or False). Standard is: conv_loop=True : and cannot be changed in the GUI.
CONV_LOOP = True
# Change some minimise opt params.
# This goes into: minimise.execute(self.min_algor, func_tol=self.opt_func_tol, max_iter=self.opt_max_iterations)
#####################
#dAuvergne_protocol.opt_func_tol = 1e-5 # Standard: opt_func_tol = 1e-25
#dAuvergne_protocol.opt_max_iterations = 1000 # Standard: opt_max_iterations = int(1e7)
#dAuvergne_protocol.opt_func_tol = 1e-10 # Standard: opt_func_tol = 1e-25
#dAuvergne_protocol.opt_max_iterations = int(1e5) # Standard: opt_max_iterations = int(1e7)
#####################################
# The results dir.
var = 'result_06'
results_dir = os.getcwd() + os.sep + var
# Save the state before running. Open and check in GUI!
state.save(state=var+'_ini.bz2', dir=results_dir, force=True)
# To check in GUI
# relax -g
# File -> Open relax state
# In folder "result_03" open "result_03_ini.bz2"
# View -> Data pipe editor
# Right click on pipe, and select "Associate with a new auto-analysis"
dAuvergne_protocol(pipe_name=name, pipe_bundle=bundle_name, results_dir=results_dir, diff_model=DIFF_MODEL, mf_models=MF_MODELS, local_tm_models=LOCAL_TM_MODELS, grid_inc=GRID_INC, min_algor=MIN_ALGOR, mc_sim_num=MC_NUM, max_iter=MAX_ITER, conv_loop=CONV_LOOP)
</source>
We use [http://www.dayid.org/comp/tm.html tmux] to make a terminal-session, we can get back to,
if our own terminal connection get closed.
* start a new session: '''tmux new -s relax06'''
* re-attach a detached session: '''tmux a -t relax06'''
Run with
<source lang="bash">
# Make terminal-session
tmux new -s relax06
relax 06_run_def_MC20_MAX_ITER20.py -t 06_run_def_MC20_MAX_ITER20.log
</source>
== To run on Haddock ==