def genetic_algorithm():
"""
Run the genetic algorithm.
"""
warnings.filterwarnings(
"ignore", ".*GUI is implemented.*"
) # filter out the plotting warning about deprecation
print('You are running the evolutionary algorithm')
#This function sets all of the values of the number of parents, children, and mutation amount
(num_init_parents, num_init_children, init_voltage, filename, num_parents,
num_children, mutation_percentage, data_acquisition_device,
mirror_communication_device, fom_num, deformable_mirror,
zernike_polynomial_mode, radial_order) = initialization_f.initialize()
print('\nNOTE: LabView must be open in order to run the program\n')
print('Here are the options you have while the program is running:')
print('\tPress the enter key (or s) to stop the program')
print('\tPress "m" if you would like to change the mutation percentage')
print('\tPress "c" if you would like to change the number of children')
print('\tPrees "p" if you would like to change the number of parents')
print('Press "r" then enter key to run the program')
while True: # run an infinite loop until a key is pressed
keyboard_input = input() # get input from the user
if keyboard_input == '\r' or keyboard_input == '\n' or keyboard_input == 'r': # if the key pressed was the enter key
break # break out of the infinite loop
opt_device = optimization_devices.initialize_opt_device(
deformable_mirror, zernike_polynomial_mode,
radial_order) # open and initialize the optimization device being used
daq_device = daq_devices.initialize_daq_device(
data_acquisition_device,
fom_num) # open and initialize the data acquisition device being used
opt_comm_device = opt_com_devices.initialize_comm_device(
mirror_communication_device
) # open and initialize the optimization communication device being used
print('Starting...')
start_time = time.time() # determine the time when the algorithm starts
iteration_number = 0
parents = people.parent_group(
num_init_parents, opt_device, init_voltage,
filename) # create parents from initialization constraints
children = people.child_group(
num_init_children, parents,
opt_device) # create children from the given parents and do crossover
children.mutate(mutation_percentage, opt_device) # mutate the children
all_people = people.person_group(
parents, children
) # combine all of the children and parents into a single object
all_people.test_and_sort_people(
opt_device, daq_device, opt_comm_device
) # measure the figures of merits and sort the people so the highest figure of merit is 0th indexed
best_person = all_people.people[
0] # the best person is the 0th indexed person in all_people
print('best_person\n', best_person.figure_of_merit
) # show the best person's genes and figure of merit
best_people = np.empty(
0, people.person) # initialize the array of best people
best_people = np.append(best_people, best_person) # store best person
past_figures_of_merit = all_people.best_figures_of_merit(num_parents)
print('Time to run: ',
time.time() - start_time, ' seconds'
) # print out the number of seconds since the algorithm started
while True: # run an infinite loop until user says to stop
if msvcrt.kbhit(): # if the keyboard was hit
keyboard_input = msvcrt.getwche() # determine what key was pressed
if keyboard_input == '\r' or keyboard_input == 's': # if the enter key was pressed
break # get out of the while loop
elif keyboard_input == 'm': # if the m key was pressed
print('\nThis is the current mutation percentage: ',
mutation_percentage)
mutation_percentage = initialization_f.change_value(
'float', 0, 100
) # change the mutation percentage to what the user wants
elif keyboard_input == 'c': # if the c key was pressed
print('\nThis is the current number of children: ',
num_children)
num_children = initialization_f.change_value(
'int', num_parents -
1) # change the number of children to what the user wants
elif keyboard_input == 'p': # if the p key was pressed
print('\nThis is the current number of parents: ', num_parents)
num_parents = initialization_f.change_value(
'float', 0, num_children +
1) # change the number of parents to what the user wants
parents = people.parent_group(
num_parents, opt_device, None, None,
all_people) # create parents from the best performing children
children = people.child_group(
num_children, parents,
opt_device) # create children from the just created parents
children.mutate(mutation_percentage, opt_device) # mutate the children
all_people = people.person_group(
parents, children
) # combine all of the children and parents into a single container
all_people.test_and_sort_people(
opt_device, daq_device, opt_comm_device
) # measure the figures of merits and sort the people so the highest figure of merit is 0th indexed
new_best_person = all_people.people[
0] # the best person is the 0th indexed person in all_people
best_people = np.append(best_people,
new_best_person) # store best person
if new_best_person.figure_of_merit > best_person.figure_of_merit: # if the current best FOM is better than the overall best FOM
best_person = new_best_person # if the new best person is better, they are the overall best person ever
print(
'best_person\n',
best_person.figure_of_merit) # print out the best person ever made
opt_device.plot_object(
new_best_person.genes, best_person.genes, iteration_number
) # plot the current best and overall best genes on the optimization device
figures_of_merit = np.concatenate(
(past_figures_of_merit,
all_people.best_figures_of_merit(num_parents)),
axis=1
) # concatenate the previous figure of merit matrix with the current figures of merit
iteration_number, past_figures_of_merit = plot_f.plot_performance(
iteration_number,
figures_of_merit) # plot the progressions of figures of merit
print('Time to run: ',
time.time() - start_time, ' seconds'
) # print out the number of seconds since the algorithm started
daq_device.shut_down() # shut off the data acquisition device
print(
'\nWhat would you like to save?'
) # once the loop has finished, the user decides what to do with the genes made
while True: # create an infinite loop
print(
"\nFor writing the best person's actuator voltages to a file, input 'write'"
)
print('For saving graph data, input "graph"')
print(
"For saving the best person from each iteration, input 'all_best'")
print("For saving everything, input 'everything'")
print('For doing nothing, input "none"')
saving_option = input() # get user input for what they want to do
directory_path = os.path.dirname(
os.path.abspath(__file__)) # get the current directory's path
if saving_option == 'write': # if they want to write the genes to a file
directory_path = save_different_directory(
opt_device.default_directory)
print(
"Enter the file name you want to be saved (do not include the file extension):\nNote: this will overwrite a file with the same name"
)
filename = input(
) # get user input from for what filename they want
opt_device.write_genes(
best_person.genes,
directory_path + filename) # write the genes to the input file
if anything_else(
) == 'n': # if they don't want to do anything else
break # break out of the while loop
elif saving_option == 'graph': # if the user wants to save the graph
directory_path = save_different_directory(directory_path +
file_f.FOM_GRAPH_FOLDER)
print(
"Enter the file name you want to be saved without the file extension (for test.csv & test.png, input test):",
"\nNote: This saves a .csv of graph's values and a pdf of the python figure\nAlso, this will overwrite a file with the same name"
)
filename = input(
) # get user input from for what filename they want
file_f.write_figures_of_merit(
figures_of_merit, directory_path + filename
) # write the figures of merit to a comma separated value file
plt.savefig(directory_path + '%s.png' % filename,
dpi=300,
bbox_inches='tight')
if anything_else(
) == 'n': # if they don't want to do anything else
break # break out of the while loop
elif saving_option == 'all_best': # if the user wants to save all of the best genes
directory_path = save_different_directory(
opt_device.default_directory)
print(
"Enter the folder you want the best people to be saved into (to have everything saved in a",
"folder named 'test', input 'test'.:\nNote: this will overwrite a folder with the same name",
"\nNote: It will save each person according to its iteration number. The ith person will be saved as 'i.[file_extension]'"
)
folder = input()
new_dir_path = directory_path + "\\" + folder + "\\" # determine the directory path to save the genes to
if not os.path.exists(
new_dir_path): # if that path doesn't exist, create it
os.makedirs(new_dir_path)
for i in range(best_people.size): # for each best person
opt_device.write_genes(best_people[i].genes, new_dir_path +
str(i)) # write each gene
if anything_else(
) == 'n': # if they don't want to do anything else
break # break out of the while loop
elif saving_option == 'everything':
print("The directory it is saving into is " + directory_path +
file_f.FOM_GRAPH_FOLDER + " and " +
opt_device.default_directory)
print(
"Enter the file name you want for all of these files (for 'test.csv', 'test.png', etc., input 'test'.\tNote: this will overwrite a file with the same name"
)
filename = input()
opt_device.write_genes(
best_person.genes, opt_device.default_directory +
filename) # write the genes to the input file
file_f.write_figures_of_merit(
figures_of_merit,
directory_path + file_f.FOM_GRAPH_FOLDER + filename
) # write the figures of merit to a comma separated value file
plt.savefig(directory_path + file_f.FOM_GRAPH_FOLDER +
'%s.png' % filename,
dpi=300,
bbox_inches='tight')
new_dir_path = opt_device.default_directory + "\\" + filename + "\\" # determine the path to save the best people
if not os.path.exists(
new_dir_path
): # if the directory path doesn't exist, create it
os.makedirs(new_dir_path)
for i in range(best_people.size): # for each best person
opt_device.write_genes(best_people[i].genes, new_dir_path +
str(i)) # write each gene
if anything_else(
) == 'n': # if they don't want to do anything else
break # break out of the while loop
elif saving_option == 'none': # if the user doesn't want to do anything with the data
break # break out of the while loop
else: # if they didn't enter one of the options
print("You didn't enter a correct command")
def genetic_algorithm():
'''Initilize IC library and open the CCD'''
# exposure = -8 #exposure time of the CCD, 2^-5 seconds
# cam = ic_capture.ccd_open(exposure)
warnings.filterwarnings("ignore",".*GUI is implemented.*") # filter out the plotting warning about deprecation
"""Start, run, and end the genetic algorithm."""
print('You are running the genetic algorithm')
#This function sets all of the values of the number of parents, children, and mutation amount
num_genes, num_init_parents, num_init_children, init_voltage, filename, num_parents, num_children, mutation_percentage, data_acquisition_device, fom_num = initialization_f.initialize()
"""Note: To change the default values, go into genetic_algorithm.ini"""
print('\nNOTE: LabView must be open in order to run the program\n')
print('Here are the options you have while the program is running:')
print('\tPress the enter key (or s) to stop the program')
print('\tPress "m" if you would like to change the mutation percentage')
print('\tPress "c" if you would like to change the number of children')
print('\tPrees "p" if you would like to change the number of parents')
print('Press "r" then enter key to run the program')
while True: # run an infinite loop until a key is pressed
keyboard_input = input() # get input from the user
if keyboard_input == '\r' or keyboard_input == '\n' or keyboard_input == 'r': # if the key pressed was the enter key
break # break out of the infinite loop
print('Starting...')
start_time = time.time() # determine the time when the algorithm starts
dm_actuators = mirror_f.actuator_array() # initialize the class to determine if actuator voltages break the mirror or not
device = data_acq_f.data_acqusition(data_acquisition_device, fom_num) # open and initialize the data acquisition device being used
iteration_number = 0
parents = people.parent_group(num_init_parents, num_genes, init_voltage, filename) # create parents from above constraints
children = people.child_group(num_init_children, parents, dm_actuators) # create children from the given parents
children.mutate(mutation_percentage, dm_actuators) # mutate the children
all_people = people.person_group(parents, children) # combine all of the children and parents into a single container
all_people.test_and_sort_people(dm_actuators, device) # measure the figures of merits and sort the people so the highest figure of merit is 0th indexed
best_person = all_people.people[0] # the best person is the 0th indexed person in all_people
print('best_person\n', best_person.figure_of_merit) # show the best person's genes and figure of merit
past_figures_of_merit = all_people.best_figures_of_merit(num_parents)
print('Time to run: ', time.time() - start_time, ' seconds') # print out the number of seconds since the algorithm started
while True: # run an infinite loop until user says to stop
if msvcrt.kbhit(): # if the keyboard was hit
keyboard_input = msvcrt.getwche() # determine what key was pressed
if keyboard_input == '\r' or keyboard_input == 's': # if the enter key was pressed
break # get out of the while loop
elif keyboard_input == 'm': # if the m key was pressed
print('\nThis is the current mutation percentage: ', mutation_percentage)
mutation_percentage = initialization_f.change_value('float', 0, 100) # change the mutation percentage to what the user wants
elif keyboard_input == 'c': # if the c key was pressed
print('\nThis is the current number of children: ', num_children)
num_children = initialization_f.change_value('int', num_parents-1) # change the number of children to what the user wants
elif keyboard_input == 'p': # if the p key was pressed
print('\nThis is the current number of parents: ', num_parents)
num_parents = initialization_f.change_value('float', 0, num_children+1) # change the number of parents to what the user wants
parents = people.parent_group(num_parents,num_genes, None, None, all_people) # create parents from the best performing children
children = people.child_group(num_children, parents, dm_actuators) # create children from the just created parents
children.mutate(mutation_percentage, dm_actuators) # mutate the children
all_people = people.person_group(parents, children) # combine all of the children and parents into a single container
all_people.test_and_sort_people(dm_actuators) # measure the figures of merits and sort the people so the highest figure of merit is 0th indexed
new_best_person = all_people.people[0] # the best person is the 0th indexed person in all_people
if new_best_person.figure_of_merit > best_person.figure_of_merit: # determine if the best person's figure of merit in this run is better than the current best person's figure of merit
best_person = new_best_person # if the new best person is better, they are the overall best person ever
print('best_person\n', best_person.figure_of_merit) # print out the best person ever made
figures_of_merit = np.concatenate((past_figures_of_merit, all_people.best_figures_of_merit(num_parents)), axis=1) # concatenate the previous figure of merit matrix with the current figures of merit
iteration_number, past_figures_of_merit = plot_f.plot_performance(iteration_number, figures_of_merit) # plot the progressions of figures of merit
print('Time to run: ', time.time() - start_time, ' seconds') # print out the number of seconds since the algorithm started
'''Close CCD and stop using c library'''
# ic_capture.ccd_close()
device.shut_down() # shut off the data acquisition device
print('What would you like to do with the best person?') # once the loop has finished, the user decides what to do with the genes made
while True: # create an infinite loop
print('\tFor writing the actuator voltages to a file, input "write"')
print('\tFor saving graph data, input "graph"')
print('\tFor doing nothing, input "none"')
saving_option = input() # get user input for what they want to do
if saving_option == 'write': # if they want to write the genes to a file
directory_path = os.path.dirname(os.path.abspath(__file__)) # get the current directory's path
print("\tThe default directory is " + directory_path + ADF_FOLDER)
directory_path = save_different_directory(directory_path + ADF_FOLDER)
print("Enter the file name you want to be saved (include the .adf extension):\nNote: this will overwrite a file with the same name")
filename = input() # get user input from for what filename they want
file_f.write_adf(best_person, directory_path + filename) # write the genes to the input file
if anything_else() == 'n': # if they don't want to do anything else
break # break out of the while loop
elif saving_option == 'graph': # if the user wants to save the graph
directory_path = os.path.dirname(os.path.abspath(__file__)) # get the current directory's path
print("\tThe default directory is " + + directory_path + FOM_GRAPH_FOLDER)
directory_path = save_different_directory(directory_path + FOM_GRAPH_FOLDER)
print("Enter the file name you want to be saved (for test.csv, input test):\nNote: this will overwrite a file with the same name")
filename = input() # get user input from for what filename they want
file_f.write_figures_of_merit(figures_of_merit, directory_path + filename) # write the figures of merit to a comma separated value file
plt.savefig( directory_path +'%s.pdf' % filename, dpi = 300, bbox_inches = 'tight')
if anything_else() == 'n': # if they don't want to do anything else
break # break out of the while loop
elif saving_option == 'none': # if the user doesn't want to do anything with the data
break # break out of the while loop
else: # if they didn't enter one of the options
print("You didn't enter a correct command")
print("Voltages would've broken mirror.")
break
if __name__ == "__main__":
while True: # create a while loop to determine if the user is in zernike polynomial mode
print("Are you using the Zernike polynomial mode for writing voltages to the mirror?")
print("Enter 'y' or 'n'")
response = input()
if response == 'y':
zernike_polynomial_mode = True
print("What radial order are you using? Enter an integer.\nThe current value is 4.")
radial_order = init_f.change_value('int', 0, 9) # determine what radial order to use
break
elif response == 'n':
zernike_polynomial_mode = False
radial_order = 0
break
else:
print("You didn't enter a 'y' or an 'n'")
# determine which optimization device the user wants to use
print('\nThese are the optimization devices you can send voltages to (only send voltages to mirrors): ')
print(optimization_devices.OPT_DEVICES)
print("Please enter the index of the device you would like to use.")
print("Indices go from 0 to", len(optimization_devices.OPT_DEVICES)-1)
while True:
index = int(input())