def test_replace_dict_values(self):
name = "something {value1} - {value2} something else {value3}"
dictionary = {
'value1': 'bla bla',
'value2': np.array([15]),
'value3': 76
}
new_name = misc.replace_dict_values(name, dictionary)
expected_new_name = 'something bla bla - [15] something else 76'
self.assertEqual(new_name, expected_new_name)
name = "something {value1} - {value2} something else {value3}"
dictionary = {
'value1': 'bla bla',
'value2': np.array([5, 10, 15, 20, 25, 30]),
'value3': 76
}
new_name = misc.replace_dict_values(name, dictionary, True)
expected_new_name = 'something bla bla - [5_(5)_30] something else 76'
self.assertEqual(new_name, expected_new_name)
# Value2 is not an arithmetic progression
dictionary2 = {
'value1': 'bla bla',
'value2': np.array([5, 10, 18, 20, 25, 30]),
'value3': 76
}
new_name2 = misc.replace_dict_values(name, dictionary2)
expected_new_name2 \
= 'something bla bla - [5,10,18,20,25,30] something else 76'
self.assertEqual(new_name2, expected_new_name2)
# Value3 has parts that are arithmetic progressions and others that
# are not an arithmetic progression
dictionary3 = {
'value1': 'bla bla',
'value2': np.array([2, 5, 10, 15, 20, 25, 30, 31, 32, 50]),
'value3': 76
}
new_name3 = misc.replace_dict_values(name, dictionary3, True)
expected_new_name3 \
= 'something bla bla - [2,5_(5)_30,31,32,50] something else 76'
self.assertEqual(new_name3, expected_new_name3)
# Now we test the same thing, but with the third parameter being False
new_name4 = misc.replace_dict_values(name, dictionary3, False)
expected_new_name4 \
= 'something bla bla - [2,5:5:30,31,32,50] something else 76'
self.assertEqual(new_name4, expected_new_name4)
def main():
"""Main function of the script.
"""
parser = argparse.ArgumentParser()
parser.add_argument('first',
help="The name of the first SimulationResults file.")
parser.add_argument('second',
help="The name of the second SimulationResults file.")
parser.add_argument('output',
help=("The name that will be used to save the combined "
"SimulationResults file."),
nargs='?')
args = parser.parse_args()
first = SimulationResults.load_from_file(args.first)
second = SimulationResults.load_from_file(args.second)
union = combine_simulation_results(first, second)
if args.output is None:
output = replace_dict_values(first.original_filename,
union.params.parameters,
filename_mode=True)
else:
output = args.output
if output == args.first or output == args.second:
raise RuntimeError(
"output filename must be different from the filename of either"
" of the two SimulationResults.")
# Finally save to the output file
union.save_to_file(output)
def main():
"""Main function of the script.
"""
parser = argparse.ArgumentParser()
parser.add_argument('first',
help="The name of the first SimulationResults file.")
parser.add_argument('second',
help="The name of the second SimulationResults file.")
parser.add_argument(
'output',
help=("The name that will be used to save the combined "
"SimulationResults file."),
nargs='?')
args = parser.parse_args()
first = SimulationResults.load_from_file(args.first)
second = SimulationResults.load_from_file(args.second)
union = combine_simulation_results(first, second)
if args.output is None:
output = replace_dict_values(first.original_filename,
union.params.parameters,
filename_mode=True)
else:
output = args.output
if output == args.first or output == args.second:
raise RuntimeError(
"output filename must be different from the filename of either"
" of the two SimulationResults.")
# Finally save to the output file
union.save_to_file(output)
def main():
"""Main function of the script.
"""
parser = argparse.ArgumentParser()
parser.add_argument('name', help="The name of the SimulationResults file.")
parser.add_argument('folder',
help="The name of the second SimulationResults file.",
nargs='?')
args = parser.parse_args()
name = args.name
folder = args.folder # This will be None, if not provided
results = SimulationResults.load_from_file(name)
original_filename = results.original_filename
# xxxxxxxxxx Remove the .pickle file extension xxxxxxxxxxxxxxxxxxxxxxxx
# If partial_filename has a 'pickle' extension we remove it from the name
original_filename_no_ext, ext = os.path.splitext(original_filename)
if ext == '.pickle':
original_filename = original_filename_no_ext
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
unpacked_params_list = results.params.get_unpacked_params_list()
all_results_names = results.get_result_names()
for i, p in enumerate(results.params.get_unpacked_params_list()):
partial_filename = get_partial_results_filename(
original_filename, p, folder)
partial_filename_with_replacements = replace_dict_values(
partial_filename, results.params.parameters, filename_mode=True)
if partial_filename_with_replacements == name:
raise RuntimeError('invalid name')
partial_param = unpacked_params_list[i]
partial_result = SimulationResults()
partial_result.set_parameters(partial_param)
for result_name in all_results_names:
partial_result.add_result(results[result_name][i])
partial_result.current_rep = results.runned_reps[i]
# Try to save the partial results. If we get an error
try:
# If we get an IOError exception here, maybe we are trying to
# save to a folder and the folder does not exist yet.
partial_result.save_to_file(partial_filename_with_replacements)
except IOError as e:
if folder is not None:
# Lets create the folder...
os.mkdir(folder)
# ... and try to save the file again.
partial_result.save_to_file(partial_filename_with_replacements)
else:
raise e
def test_replace_dict_values(self):
name = "something {value1} - {value2} something else {value3}"
dictionary = {'value1': 'bla bla',
'value2': np.array([15]),
'value3': 76}
new_name = misc.replace_dict_values(name, dictionary)
expected_new_name = 'something bla bla - [15] something else 76'
self.assertEqual(new_name, expected_new_name)
name = "something {value1} - {value2} something else {value3}"
dictionary = {'value1': 'bla bla',
'value2': np.array([5, 10, 15, 20, 25, 30]),
'value3': 76}
new_name = misc.replace_dict_values(name, dictionary, True)
expected_new_name = 'something bla bla - [5_(5)_30] something else 76'
self.assertEqual(new_name, expected_new_name)
# Value2 is not an arithmetic progression
dictionary2 = {'value1': 'bla bla',
'value2': np.array([5, 10, 18, 20, 25, 30]),
'value3': 76}
new_name2 = misc.replace_dict_values(name, dictionary2)
expected_new_name2 \
= 'something bla bla - [5,10,18,20,25,30] something else 76'
self.assertEqual(new_name2, expected_new_name2)
# Value3 has parts that are arithmetic progressions and others that
# are not an arithmetic progression
dictionary3 = {'value1': 'bla bla',
'value2': np.array([2, 5, 10, 15, 20, 25,
30, 31, 32, 50]),
'value3': 76}
new_name3 = misc.replace_dict_values(name, dictionary3, True)
expected_new_name3 \
= 'something bla bla - [2,5_(5)_30,31,32,50] something else 76'
self.assertEqual(new_name3, expected_new_name3)
# Now we test the same thing, but with the third parameter being False
new_name4 = misc.replace_dict_values(name, dictionary3, False)
expected_new_name4 \
= 'something bla bla - [2,5:5:30,31,32,50] something else 76'
self.assertEqual(new_name4, expected_new_name4)
def get_result_from_file():
"""
Load the SimulationResults object from the file and return it.
"""
config_file = 'greedy_config_file.txt'
# xxxxxxxxxx Config spec for the config file xxxxxxxxxxxxxxxxxxxxxxxxxx
spec = """[Grid]
cell_radius=float(min=0.01, default=1.0)
num_cells=integer(min=3,default=3)
num_clusters=integer(min=1,default=1)
[Scenario]
NSymbs=integer(min=10, max=1000000, default=200)
SNR=real_numpy_array(min=-50, max=100, default=0:5:31)
M=integer(min=4, max=512, default=4)
modulator=option('QPSK', 'PSK', 'QAM', 'BPSK', default="PSK")
Nr=integer_scalar_or_integer_numpy_array_check(min=2,default=3)
Nt=integer_scalar_or_integer_numpy_array_check(min=2,default=3)
Ns=integer_scalar_or_integer_numpy_array_check(min=1,default=3)
N0=float(default=-116.4)
scenario=string_list(default=list('Random', 'NoPathLoss'))
[IA Algorithm]
max_iterations=integer(min=1, default=120)
initialize_with=string_list(default=list('random'))
stream_sel_method=string_list(default=list('greedy', 'brute'))
[General]
rep_max=integer(min=1, default=2000)
max_bit_errors=integer(min=1, default=3000)
unpacked_parameters=string_list(default=list('SNR','stream_sel_method','scenario','initialize_with'))
""".split("\n")
params = SimulationParameters.load_from_config_file(config_file, spec)
# xxxxx Results base name xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
base_name = ("IA_stream_sel_results_{SNR}_{M}-{modulator}_{Nr}x{Nt}_({Ns})"
"_MaxIter_{max_iterations}_({initialize_with})")
base_name = misc.replace_dict_values(base_name, params.parameters, True)
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
# xxxxxxxxxx Get the SimulationResults objects xxxxxxxxxxxxxxxxxxxxxxxx
results = SimulationResults.load_from_file(
'greedy_{0}.pickle'.format(base_name))
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
return results
def get_result_from_file():
"""
Load the SimulationResults object from the file and return it.
"""
config_file = 'greedy_config_file.txt'
# xxxxxxxxxx Config spec for the config file xxxxxxxxxxxxxxxxxxxxxxxxxx
spec = """[Grid]
cell_radius=float(min=0.01, default=1.0)
num_cells=integer(min=3,default=3)
num_clusters=integer(min=1,default=1)
[Scenario]
NSymbs=integer(min=10, max=1000000, default=200)
SNR=real_numpy_array(min=-50, max=100, default=0:5:31)
M=integer(min=4, max=512, default=4)
modulator=option('QPSK', 'PSK', 'QAM', 'BPSK', default="PSK")
Nr=integer_scalar_or_integer_numpy_array_check(min=2,default=3)
Nt=integer_scalar_or_integer_numpy_array_check(min=2,default=3)
Ns=integer_scalar_or_integer_numpy_array_check(min=1,default=3)
N0=float(default=-116.4)
scenario=string_list(default=list('Random', 'NoPathLoss'))
[IA Algorithm]
max_iterations=integer(min=1, default=120)
initialize_with=string_list(default=list('random'))
stream_sel_method=string_list(default=list('greedy', 'brute'))
[General]
rep_max=integer(min=1, default=2000)
max_bit_errors=integer(min=1, default=3000)
unpacked_parameters=string_list(default=list('SNR','stream_sel_method','scenario','initialize_with'))
""".split("\n")
params = SimulationParameters.load_from_config_file(config_file, spec)
# xxxxx Results base name xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
base_name = ("IA_stream_sel_results_{SNR}_{M}-{modulator}_{Nr}x{Nt}_({Ns})"
"_MaxIter_{max_iterations}_({initialize_with})")
base_name = misc.replace_dict_values(base_name, params.parameters, True)
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
# xxxxxxxxxx Get the SimulationResults objects xxxxxxxxxxxxxxxxxxxxxxxx
results = SimulationResults.load_from_file(
'greedy_{0}.pickle'.format(base_name))
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
return results
def main_plot(index=0): # pylint: disable=R0914,R0915
"""
Function called to plot the results from a previous simulation.
"""
from matplotlib import pyplot as plt
config_file = "greedy_config_file.txt"
# xxxxxxxxxx Config spec for the config file xxxxxxxxxxxxxxxxxxxxxxxxxx
spec = """[Grid]
cell_radius=float(min=0.01, default=1.0)
num_cells=integer(min=3,default=3)
num_clusters=integer(min=1,default=1)
[Scenario]
NSymbs=integer(min=10, max=1000000, default=200)
SNR=real_numpy_array(min=-50, max=100, default=0:5:31)
M=integer(min=4, max=512, default=4)
modulator=option('QPSK', 'PSK', 'QAM', 'BPSK', default="PSK")
Nr=integer_scalar_or_integer_numpy_array_check(min=2,default=3)
Nt=integer_scalar_or_integer_numpy_array_check(min=2,default=3)
Ns=integer_scalar_or_integer_numpy_array_check(min=1,default=3)
N0=float(default=-116.4)
scenario=string_list(default=list('Random', 'NoPathLoss'))
[IA Algorithm]
max_iterations=integer(min=1, default=120)
initialize_with=string_list(default=list('random'))
stream_sel_method=string_list(default=list('greedy', 'brute'))
[General]
rep_max=integer(min=1, default=2000)
max_bit_errors=integer(min=1, default=3000)
unpacked_parameters=string_list(default=list('SNR','stream_sel_method','scenario','initialize_with'))
[Plot]
max_iterations_plot=integer(default=5)
initialize_with_plot=option('random', 'alt_min', default='random')
""".split(
"\n"
)
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
# xxxxx Parameters xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
params = SimulationParameters.load_from_config_file(config_file, spec)
max_iterations = params["max_iterations_plot"]
# initialize_with = params['initialize_with_plot']
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
# xxxxx Results base name xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
base_name = (
"IA_stream_sel_results_{SNR}_{M}-{modulator}_{Nr}x{Nt}" "_({Ns})_MaxIter_{max_iterations}_({initialize_with})"
)
# base_name = ("results_{SNR}_{M}-{modulator}_{Nr}x{Nt}_({Ns})_MaxIter"
# "_{max_iterations}_{initialize_with}")
base_name = misc.replace_dict_values(base_name, params.parameters, True)
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
fig, ax = plt.subplots(nrows=1, ncols=1)
fig2, ax2 = plt.subplots(nrows=1, ncols=1)
# Include results with the Greedy and Brute Force stream selection
# algorithms
results = SimulationResults.load_from_file("greedy_{0}.pickle".format(base_name))
# We only get the parameters from the greedy object, since we use the
# same parameters for greedy and brute force
parameters_dict = results.params.parameters
# Fixed parameters for the Greedy stream selection algorithm in the
# "NoPathLoss" scenario
greedy_nopl_fixed_params = {
"max_iterations": max_iterations,
"stream_sel_method": "greedy",
"initialize_with": "random",
"scenario": "NoPathLoss",
}
# Fixed parameters for the Greedy stream selection algorithm in the
# "Random" scenario (path loss with random positions for the users)
greedy_random_fixed_params = {
"max_iterations": max_iterations,
"stream_sel_method": "greedy",
"initialize_with": "random",
"scenario": "Random",
}
# Fixed parameters for the Brute Force stream selection algorithm in
# the "NoPathLoss" scenario
brute_nopl_fixed_params = {
"max_iterations": max_iterations,
"stream_sel_method": "brute",
"initialize_with": "random",
"scenario": "NoPathLoss",
}
# Fixed parameters for the Brute Force stream selection algorithm in
# the "Random" scenario (path loss with random positions for the users)
brute_random_fixed_params = {
"max_iterations": max_iterations,
"stream_sel_method": "brute",
"initialize_with": "random",
"scenario": "Random",
}
_plot_ber(results, greedy_nopl_fixed_params, ax, "Greedy (No PL)", "-b*")
_plot_ber(results, greedy_random_fixed_params, ax, "Greedy (With PL)", "-r*")
_plot_sum_capacity(results, greedy_nopl_fixed_params, ax2, "Greedy (No PL)", "-b*")
_plot_sum_capacity(results, greedy_random_fixed_params, ax2, "Greedy (With PL)", "-r*")
_plot_ber(results, brute_nopl_fixed_params, ax, "Brute Force (No PL)", "-co")
_plot_ber(results, brute_random_fixed_params, ax, "Brute Force (With PL)", "-mo")
_plot_sum_capacity(results, brute_nopl_fixed_params, ax2, "Brute Force (No PL)", "-co")
_plot_sum_capacity(results, brute_random_fixed_params, ax2, "Brute Force (With PL)", "-mo")
# Get a dataframe with the results in a well organized way
results_dataframe = get_dataframe_from_results(results)
filename = "greedy_results_{M}-{modulator}.csv".format(**results.params.parameters)
results_dataframe.to_csv(filename, index_label="SNR")
# xxxxxxxxxx BER Plot Options xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
ax.set_xlabel("SNR")
ax.set_ylabel("BER")
title = (
"BER for Different Algorithms ({max_iterations} Max Iterations)\n"
"K={num_cells}, Nr={Nr}, Nt={Nt}, Ns={Ns}, {M}-{modulator}"
)
title = title.replace("{max_iterations}", str(max_iterations))
ax.set_title(title.format(**parameters_dict))
ax.set_yscale("log")
ax.legend(fancybox=True, shadow=True, loc="best")
ax.grid(True, which="both", axis="both")
# plt.show(block=False)
fig_base_name = "{Nr}x{Nt}_({Ns})_{M}-{modulator}".format(**params.parameters)
fig.savefig("ber_{0}.pdf".format(fig_base_name))
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
# xxxxxxxxxx Sum Capacity Plot Options xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
ax2.set_xlabel("SNR")
ax2.set_ylabel("Sum Capacity")
title = (
"Sum Capacity for Different Algorithms ({max_iterations} Max "
"Iterations)\nK={num_cells}, Nr={Nr}, Nt={Nt}, Ns={Ns}, {M}-"
"{modulator}"
)
title = title.replace("{max_iterations}", str(max_iterations))
ax2.set_title(title.format(**parameters_dict))
ax2.legend(fancybox=True, shadow=True, loc=2)
ax2.grid(True, which="both", axis="both")
# plt.show()
fig2.savefig("sum_capacity_{0}.pdf".format(fig_base_name))
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
plt.show()
def main_plot(index=0): # pylint: disable=R0914,R0915
"""
Function called to plot the results from a previous simulation.
Parameters
----------
index : int
"""
from matplotlib import pyplot as plt
config_file = 'greedy_config_file.txt'
# xxxxxxxxxx Config spec for the config file xxxxxxxxxxxxxxxxxxxxxxxxxx
spec = """[Grid]
cell_radius=float(min=0.01, default=1.0)
num_cells=integer(min=3,default=3)
num_clusters=integer(min=1,default=1)
[Scenario]
NSymbs=integer(min=10, max=1000000, default=200)
SNR=real_numpy_array(min=-50, max=100, default=0:5:31)
M=integer(min=4, max=512, default=4)
modulator=option('QPSK', 'PSK', 'QAM', 'BPSK', default="PSK")
Nr=integer_scalar_or_integer_numpy_array_check(min=2,default=3)
Nt=integer_scalar_or_integer_numpy_array_check(min=2,default=3)
Ns=integer_scalar_or_integer_numpy_array_check(min=1,default=3)
N0=float(default=-116.4)
scenario=string_list(default=list('Random', 'NoPathLoss'))
[IA Algorithm]
max_iterations=integer(min=1, default=120)
initialize_with=string_list(default=list('random'))
stream_sel_method=string_list(default=list('greedy', 'brute'))
[General]
rep_max=integer(min=1, default=2000)
max_bit_errors=integer(min=1, default=3000)
unpacked_parameters=string_list(default=list('SNR','stream_sel_method','scenario','initialize_with'))
[Plot]
max_iterations_plot=integer(default=5)
initialize_with_plot=option('random', 'alt_min', default='random')
""".split("\n")
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
# xxxxx Parameters xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
params = SimulationParameters.load_from_config_file(config_file, spec)
max_iterations = params['max_iterations_plot']
# initialize_with = params['initialize_with_plot']
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
# xxxxx Results base name xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
base_name = ("IA_stream_sel_results_{SNR}_{M}-{modulator}_{Nr}x{Nt}"
"_({Ns})_MaxIter_{max_iterations}_({initialize_with})")
# base_name = ("results_{SNR}_{M}-{modulator}_{Nr}x{Nt}_({Ns})_MaxIter"
# "_{max_iterations}_{initialize_with}")
base_name = misc.replace_dict_values(base_name, params.parameters, True)
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
fig, ax = plt.subplots(nrows=1, ncols=1)
fig2, ax2 = plt.subplots(nrows=1, ncols=1)
# Include results with the Greedy and Brute Force stream selection
# algorithms
results = SimulationResults.load_from_file(
'greedy_{0}.pickle'.format(base_name))
# We only get the parameters from the greedy object, since we use the
# same parameters for greedy and brute force
parameters_dict = results.params.parameters
# Fixed parameters for the Greedy stream selection algorithm in the
# "NoPathLoss" scenario
greedy_nopl_fixed_params = {
'max_iterations': max_iterations,
'stream_sel_method': 'greedy',
'initialize_with': 'random',
'scenario': 'NoPathLoss'
}
# Fixed parameters for the Greedy stream selection algorithm in the
# "Random" scenario (path loss with random positions for the users)
greedy_random_fixed_params = {
'max_iterations': max_iterations,
'stream_sel_method': 'greedy',
'initialize_with': 'random',
'scenario': 'Random'
}
# Fixed parameters for the Brute Force stream selection algorithm in
# the "NoPathLoss" scenario
brute_nopl_fixed_params = {
'max_iterations': max_iterations,
'stream_sel_method': 'brute',
'initialize_with': 'random',
'scenario': 'NoPathLoss'
}
# Fixed parameters for the Brute Force stream selection algorithm in
# the "Random" scenario (path loss with random positions for the users)
brute_random_fixed_params = {
'max_iterations': max_iterations,
'stream_sel_method': 'brute',
'initialize_with': 'random',
'scenario': 'Random'
}
_plot_ber(results, greedy_nopl_fixed_params, ax, 'Greedy (No PL)', '-b*')
_plot_ber(results, greedy_random_fixed_params, ax, 'Greedy (With PL)',
'-r*')
_plot_sum_capacity(results, greedy_nopl_fixed_params, ax2,
'Greedy (No PL)', '-b*')
_plot_sum_capacity(results, greedy_random_fixed_params, ax2,
'Greedy (With PL)', '-r*')
_plot_ber(results, brute_nopl_fixed_params, ax, 'Brute Force (No PL)',
'-co')
_plot_ber(results, brute_random_fixed_params, ax, 'Brute Force (With PL)',
'-mo')
_plot_sum_capacity(results, brute_nopl_fixed_params, ax2,
'Brute Force (No PL)', '-co')
_plot_sum_capacity(results, brute_random_fixed_params, ax2,
'Brute Force (With PL)', '-mo')
# Get a dataframe with the results in a well organized way
results_dataframe = get_dataframe_from_results(results)
filename = 'greedy_results_{M}-{modulator}.csv'.format(
**results.params.parameters)
results_dataframe.to_csv(filename, index_label='SNR')
# xxxxxxxxxx BER Plot Options xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
ax.set_xlabel('SNR')
ax.set_ylabel('BER')
title = ("BER for Different Algorithms ({max_iterations} Max Iterations)\n"
"K={num_cells}, Nr={Nr}, Nt={Nt}, Ns={Ns}, {M}-{modulator}")
title = title.replace("{max_iterations}", str(max_iterations))
ax.set_title(title.format(**parameters_dict))
ax.set_yscale('log')
ax.legend(fancybox=True, shadow=True, loc='best')
ax.grid(True, which='both', axis='both')
# plt.show(block=False)
fig_base_name = "{Nr}x{Nt}_({Ns})_{M}-{modulator}".format(
**params.parameters)
fig.savefig('ber_{0}.pdf'.format(fig_base_name))
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
# xxxxxxxxxx Sum Capacity Plot Options xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
ax2.set_xlabel('SNR')
ax2.set_ylabel('Sum Capacity')
title = ("Sum Capacity for Different Algorithms ({max_iterations} Max "
"Iterations)\nK={num_cells}, Nr={Nr}, Nt={Nt}, Ns={Ns}, {M}-"
"{modulator}")
title = title.replace("{max_iterations}", str(max_iterations))
ax2.set_title(title.format(**parameters_dict))
ax2.legend(fancybox=True, shadow=True, loc=2)
ax2.grid(True, which='both', axis='both')
# plt.show()
fig2.savefig('sum_capacity_{0}.pdf'.format(fig_base_name))
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
plt.show()
def main_plot(algorithms_to_simulate, index=0): # pylint: disable=R0914,R0915
"""
Function called to plot the results from a previous simulation.
Parameters
----------
algorithms_to_simulate : list[str]
List of algorithm names to simulate.
index : int
The index to simulate.
"""
from matplotlib import pyplot as plt
if args.config is None:
config_file = 'ia_config_file.txt'
else:
config_file = args.config
# xxxxxxxxxx Config spec for the config file xxxxxxxxxxxxxxxxxxxxxxxxxx
spec = """[Scenario]
SNR=real_numpy_array(min=-50, max=100, default=0:5:31)
M=integer(min=4, max=512, default=4)
modulator=option('QPSK', 'PSK', 'QAM', 'BPSK', default="PSK")
NSymbs=integer(min=10, max=1000000, default=200)
K=integer(min=2,default=3)
Nr=integer_scalar_or_integer_numpy_array_check(min=2,default=2)
Nt=integer_scalar_or_integer_numpy_array_check(min=2,default=2)
Ns=integer_scalar_or_integer_numpy_array_check(min=1,default=1)
[IA Algorithm]
max_iterations=integer_numpy_array(min=1, default=60)
[General]
rep_max=integer(min=1, default=2000)
max_bit_errors=integer(min=1, default=3000)
unpacked_parameters=string_list(default=list('SNR'))
[Plot]
max_iterations_plot=integer(default=5)
initialize_with_plot=option('random', 'alt_min', default='random')
""".split("\n")
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
# xxxxx Parameters xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
params = SimulationParameters.load_from_config_file(config_file, spec)
max_iterations = params['max_iterations_plot']
initialize_with = params['initialize_with_plot']
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
# xxxxx Results base name xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
# Base name for all IA algorithms (except the Closed Form)
base_name = ("results_{M}-{modulator}_{Nr}x{Nt}_({Ns})_MaxIter"
"_{max_iterations}")
base_name = misc.replace_dict_values(base_name, params.parameters)
base_name2 = ("results_{M}-{modulator}_{Nr}x{Nt}_({Ns})_MaxIter"
"_{max_iterations}_{initialize_with}")
base_name2 = misc.replace_dict_values(base_name2, params.parameters)
# Base name for the closed form IA algorithm.
base_name_no_iter = base_name
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
fig, ax = plt.subplots(nrows=1, ncols=1)
fig2, ax2 = plt.subplots(nrows=1, ncols=1)
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
if 'Alt Min' in algorithms_to_simulate:
alt_min_results = SimulationResults.load_from_file(
'ia_alt_min_{0}.pickle'.format(base_name))
parameters_dict = alt_min_results.params.parameters
fixed_params = {'max_iterations': max_iterations}
_plot_ber(alt_min_results, fixed_params, ax, 'Alt. Min.', '-r*')
_plot_sum_capacity(
alt_min_results, fixed_params, ax2, 'Alt. Min.', '-r*')
if "Closed Form" in algorithms_to_simulate:
closed_form_results = SimulationResults.load_from_file(
'ia_closed_form_{0}.pickle'.format(base_name_no_iter))
parameters_dict = closed_form_results.params.parameters
fixed_params = {}
_plot_ber(closed_form_results, fixed_params, ax, 'Closed Form', '-b*')
_plot_sum_capacity(
closed_form_results, fixed_params, ax2, 'Closed Form', '-b*')
if "Max SINR" in algorithms_to_simulate:
max_sinrn_results = SimulationResults.load_from_file(
'ia_max_sinr_{0}.pickle'.format(base_name2))
parameters_dict = max_sinrn_results.params.parameters
fixed_params = {'max_iterations': max_iterations,
'initialize_with': initialize_with}
_plot_ber(max_sinrn_results, fixed_params, ax, 'Max SINR', '-g*')
_plot_sum_capacity(
max_sinrn_results, fixed_params, ax2, 'Max SINR', '-g*')
if "MMSE" in algorithms_to_simulate:
mmse_results = SimulationResults.load_from_file(
'ia_mmse_{0}.pickle'.format(base_name2))
parameters_dict = mmse_results.params.parameters
fixed_params = {'max_iterations': max_iterations,
'initialize_with': initialize_with}
_plot_ber(mmse_results, fixed_params, ax, 'MMSE', '-m*')
_plot_sum_capacity(mmse_results, fixed_params, ax2, 'MMSE', '-m*')
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
# xxxxxxxxxx BER Plot Options xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
ax.set_xlabel('SNR')
ax.set_ylabel('BER')
title = ("BER for Different Algorithms ({max_iterations} Max Iterations)\n"
"K={K}, Nr={Nr}, Nt={Nt}, Ns={Ns}, {M}-{modulator}")
title = title.replace("{max_iterations}", str(max_iterations))
# noinspection PyUnboundLocalVariable
ax.set_title(title.format(**parameters_dict))
ax.set_yscale('log')
ax.legend(fancybox=True, shadow=True, loc='best')
ax.grid(True, which='both', axis='both')
# plt.show(block=False)
fig.savefig('ber_all_ia_algorithms.pgf')
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
# xxxxxxxxxx Sum Capacity Plot Options xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
ax2.set_xlabel('SNR')
ax2.set_ylabel('Sum Capacity')
title = ("Sum Capacity for Different Algorithms ({max_iterations} Max "
"Iterations)\nK={K}, Nr={Nr}, Nt={Nt}, Ns={Ns}, {M}-{modulator}")
title = title.replace("{max_iterations}", str(max_iterations))
ax2.set_title(title.format(**parameters_dict))
ax2.legend(fancybox=True, shadow=True, loc=2)
ax2.grid(True, which='both', axis='both')
# plt.show()
fig2.savefig('sum_capacity_all_ia_algorithms.pgf')
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
plt.show()
def main():
"""Main function of the script.
"""
parser = argparse.ArgumentParser()
parser.add_argument('name', help="The name of the SimulationResults file.")
parser.add_argument('folder',
help="The name of the second SimulationResults file.",
nargs='?')
args = parser.parse_args()
name = args.name
folder = args.folder # This will be None, if not provided
results = SimulationResults.load_from_file(name)
original_filename = results.original_filename
# xxxxxxxxxx Remove the .pickle file extension xxxxxxxxxxxxxxxxxxxxxxxx
# If partial_filename has a 'pickle' extension we remove it from the name
original_filename_no_ext, ext = os.path.splitext(original_filename)
if ext == '.pickle':
original_filename = original_filename_no_ext
# xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx
unpacked_params_list = results.params.get_unpacked_params_list()
all_results_names = results.get_result_names()
for i, p in enumerate(results.params.get_unpacked_params_list()):
partial_filename = get_partial_results_filename(
original_filename, p, folder)
partial_filename_with_replacements = replace_dict_values(
partial_filename,
results.params.parameters,
filename_mode=True)
if partial_filename_with_replacements == name:
raise RuntimeError('invalid name')
partial_param = unpacked_params_list[i]
partial_result = SimulationResults()
partial_result.set_parameters(partial_param)
for result_name in all_results_names:
partial_result.add_result(results[result_name][i])
partial_result.current_rep = results.runned_reps[i]
# Try to save the partial results. If we get an error
try:
# If we get an IOError exception here, maybe we are trying to
# save to a folder and the folder does not exist yet.
partial_result.save_to_file(partial_filename_with_replacements)
except IOError as e:
if folder is not None:
# Lets create the folder...
os.mkdir(folder)
# ... and try to save the file again.
partial_result.save_to_file(partial_filename_with_replacements)
else:
raise e
