else:
plt.xlabel('Training-testing split factor')
plt.ylabel('Total cost')
for values in param_values:
if values <= 10 and values >=.1:
print(values)
if sensitivity is 'r':
configs.r = float(values)
else:
if values==2 or values==3:
configs.th = int(values)
else:
configs.th = float(values)
ret = base.projective(configs.system)
if ret is not None:
size_result,success,p_value,cost_result,opt_size_result_ff = ret
sens_data_ff[values] = cost_result,success,opt_size_result_ff
configs.tway = 2
ret = twaysample.sample(configs.system)
if ret is not None:
size_result,success,p_value,cost_result,opt_size_result_2w = ret
sens_data_2w[values] = cost_result,success,opt_size_result_2w
configs.tway = 3
ret = twaysample.sample(configs.system)
if ret is not None:
size_result,success,p_value,cost_result,opt_size_result_3w = twaysample.sample(configs.system)
sens_data_3w[values] = cost_result,success,opt_size_result_3w
def run():
configs.print_detail = False
configs.plot = False
configs.fix_test_set = True
configs.fix_test_ratio = 3
configs.smooth = True
configs.transform_lambda = False
configs.add_origin_to_lambda = True
configs.extend_lambda = False
configs.projective_feature_threshold = 5
configs.repeat = configs.repeat
configs.min_corr = 0
configs.curve_selection = 'static'
if configs.show_box_pl is False:
out_file = open(configs.base_dir_out+"_result_transf_"+str(configs.transform_lambda)+'_smooth_'+str(configs.smooth)+'_'+str(configs.repeat)+'_rest','w')
out_file.truncate()
configs.chi_sq_with_random = True
for system_key in configs.all_systems:
boxpl.setup(system_key)
for i in range(3,7):
configs.projective_feature_threshold = i
print(str(system_key)+"-feature-frequencies-"+str(i)+": ")
if configs.show_box_pl is False:
out_file.write(str(system_key)+"-feature-frequencies-"+str(i)+": ")
configs.system = system_key
size_result,success,p_value,cost_result,opt_size_result = base.projective(system_key)
if p_value is not None:
if configs.show_box_pl is False:
print(str(p_value))
out_file.write(str(p_value))
else:
if configs.show_box_pl is False:
print('None')
out_file.write('None')
if configs.show_box_pl is False:
print()
out_file.write('\n')
out_file.write('Cost Result : ')
out_file.write(str(cost_result))
out_file.write('\n')
out_file.write('Size : ')
out_file.write(str(size_result))
out_file.write('\n')
out_file.write('Opt Size : ')
out_file.write(str(opt_size_result))
print()
print('Cost Result : ',str(cost_result))
print()
print()
out_file.write('\n')
out_file.write('Success rate : ')
out_file.write(str(success))
print()
print('Success rate : ',str(success))
print()
out_file.write('\n')
else:
i_ind = 'ff-'+str(i)
boxpl.data[i_ind] = cost_result
boxpl.lambda_size[i_ind] = size_result[2]
configs.tway = 2
print(str(system_key)+"-2way: ")
if configs.show_box_pl is False:
out_file.write('\n')
out_file.write(str(system_key)+"-2way: ")
size_result,success,p_value,cost_result,opt_size_result = twaysample.sample(system_key)
if p_value is not None:
if configs.show_box_pl is False:
print(str(p_value))
out_file.write(str(p_value))
else:
if configs.show_box_pl is False:
print('None')
out_file.write('None')
if configs.show_box_pl is False:
print()
out_file.write('\n')
out_file.write('Cost Result : ')
out_file.write(str(cost_result))
print()
print('Cost Result : ',str(cost_result))
print()
print()
out_file.write('\n')
out_file.write('Size : ')
out_file.write(str(size_result))
out_file.write('\n')
out_file.write('Opt Size : ')
out_file.write(str(opt_size_result))
out_file.write('\n')
out_file.write('Success rate : ')
out_file.write(str(success))
print()
print('Success rate : ',str(success))
print()
out_file.write('\n')
out_file.write('\n')
else:
boxpl.data['2-way'] = cost_result
boxpl.lambda_size['2-way'] = size_result[2]
configs.tway = 3
print(str(system_key)+"-3way: ")
if configs.show_box_pl is False:
out_file.write(str(system_key)+"-3way: ")
size_result,success,p_value,cost_result,opt_size_result = twaysample.sample(system_key)
if p_value is not None:
if configs.show_box_pl is False:
print(str(p_value))
out_file.write(str(p_value))
else:
if configs.show_box_pl is False:
print('None')
out_file.write('None')
if configs.show_box_pl is False:
print()
out_file.write('\n')
out_file.write('Cost Result : ')
out_file.write(str(cost_result))
out_file.write('\n')
out_file.write('Size : ')
out_file.write(str(size_result))
out_file.write('\n')
out_file.write('Opt Size : ')
out_file.write(str(opt_size_result))
print()
print('Cost Result : ',str(cost_result))
print()
out_file.write('\n')
print()
out_file.write('Success rate : ')
out_file.write(str(success))
print()
print('Success rate : ',str(success))
print()
out_file.write('\n')
else:
boxpl.data['3-way'] = cost_result
boxpl.lambda_size['3-way'] = size_result[2]
data_list,opt_cost,real_cost = base.progressive(system_key)
boxpl.global_min_cost = real_cost
boxpl.progressive_cost = opt_cost
if configs.show_box_pl is False:
out_file.write('\n')
print()
out_file.write('Actual min cost : ')
out_file.write(str(real_cost))
print()
print('Actual min cost : ',str(real_cost))
print()
out_file.write('\n')
print()
out_file.write('Progressive min cost : ')
out_file.write(str(opt_cost))
print()
print('Progressive min cost: ',str(opt_cost))
print()
print("---------------------------------------------")
out_file.write('\n')
out_file.write("---------------------------------------------")
out_file.write('\n')
else:
boxpl.show()