def get_number_of_components(filename, window_size, min_weight, influence_graph_grep='ig\:#', **kargs):
pos_callback = lambda x: SwarmAnalyzer.get_number_of_components_of_graph(
x, min_weight=min_weight * 2 * window_size, pre_callback=Callback.to_symmetric)
all_graph_matrices, _ = SwarmParser.read_file_and_measures(
filename, influence_graph_grep=influence_graph_grep, pos_callback=pos_callback,
window_size=window_size, **kargs)
return all_graph_matrices
def create_influence_graph(filename, influence_graph_grep='ig\:#', window_size=1000, calculate_on=1000):
pre_callback = Callback.to_symmetric
graph, _ = SwarmParser.read_file_and_measures(
filename, influence_graph_grep=influence_graph_grep, window_size=window_size, pre_callback=pre_callback,
calculate_on=calculate_on)
igraph_graph = InfluenceGraph.create_graph_from_matrix(graph[0][1])
return igraph_graph
def get_swarm_informations_from_file(filename,
informations_grep,
information_map=float,
**kargs):
_, informations = SwarmParser.read_files_and_measures(
[('', filename)],
informations_grep=informations_grep,
information_map=information_map,
**kargs)
informations = informations[''][-1] # there is no window here!
# let's get the longest information sequence
max_information_length = float("-inf")
max_information_key = None
for information_grep in informations:
if len(informations[information_grep]) > max_information_length:
max_information_length = len(informations[information_grep])
max_information_key = information_grep
iterations = [
iteration for (iteration, _) in informations[max_information_key]
]
iterations.sort()
df = pd.DataFrame({'x': iterations})
for information_grep in informations:
dict_information = dict(informations[information_grep])
df[information_grep] = [
dict_information[i] if i in dict_information else float("nan")
for i in iterations
]
return df
def get_graph_matrices_from_files(filenames,
influence_graph_grep='ig\:#',
**kargs):
pos_callback = Callback.to_symmetric
graph_matrices, _ = SwarmParser.read_files_and_measures(
filenames,
influence_graph_grep=influence_graph_grep,
pos_callback=pos_callback,
**kargs)
return graph_matrices
def plot_boxplot_fitness(sets_of_filenames,
output_filename=None,
info_grep="it\:#",
yticks_args=None,
legends=None,
showmeans=False,
showfliers=True,
whis=1.,
widths=0.7,
bootstrap=2000,
jump_lines=None,
**kargs):
"""
:param sets_of_filenames:
:param output_filename:
:param info_grep:
:return:
"""
get_infos = lambda x: [
SwarmParser.read_file_and_measures(
filename, informations_grep=info_grep, jump_lines=jump_lines)[
1][info_grep][-1][1] for filename in x
]
values = map(get_infos, sets_of_filenames)
boxes_kargs = {
'color': 'black',
'linewidth': 1.3,
'zorder': 3,
'fillstyle': 'full',
'facecolor': '#a6bddb'
}
means_kargs = {
'color': 'black',
'fillstyle': 'full',
'markerfacecolor': "black",
'marker': "s",
'markersize': 3,
'mew': 1,
'mec': 'black',
'zorder': 5
}
fliers_kargs = {
'color': 'black',
'marker': "s",
'markersize': 1,
'mew': 1.2,
'mec': 'black'
}
whiskers_kargs = {
'color': 'black',
'linewidth': 1.2,
'zorder': 2,
'linestyle': '-',
'alpha': 1.0,
'mec': 'black'
}
medians_kargs = {
'color': 'black',
'linewidth': 1.6,
'zorder': 5,
'alpha': 0.3
}
caps_kargs = {'linewidth': 1.5, 'color': 'black'}
xticks_args = None
if legends is not None:
xticks_args = (range(1, len(legends) + 1), legends)
Plotter.plot_boxplots(values,
grid_only='y',
xlim=(0.3, len(sets_of_filenames) + 0.6),
yticks_args=yticks_args,
whis=whis,
xticks_args=xticks_args,
grid=False,
widths=widths,
bootstrap=bootstrap,
boxes_kargs=boxes_kargs,
showmeans=showmeans,
showfliers=showfliers,
fliers_kargs=fliers_kargs,
means_kargs=means_kargs,
whiskers_kargs=whiskers_kargs,
medians_kargs=medians_kargs,
caps_kargs=caps_kargs,
output=output_filename,
**kargs)