def get_info_on_events(fault_tree):
events = []
event_dictionary = {
'name': get_object_name(fault_tree.top_event.name),
'mtbf': fault_tree.top_event.MTTF,
'mtbr': fault_tree.top_event.MTTR,
'reliability': fault_tree.top_event.reliability_distribution,
'maintainability': fault_tree.top_event.maintainability_distribution,
'oper_avail': fault_tree.top_event.availability_operational
}
events.append(event_dictionary)
for basic_event in fault_tree.get_basic_events():
event_dictionary = {
'name': get_object_name(basic_event.name),
'mtbf': basic_event.MTTF,
'mtbr': basic_event.MTTR,
'reliability': basic_event.reliability_distribution,
'maintainability': basic_event.maintainability_distribution,
'oper_avail': basic_event.availability_operational
}
events.append(event_dictionary)
return events
def plot_arbitrary_distribution_compare(name, metric, times, linspace,
theoretical):
# Reliability for now
# Maybe fix inconsistencies with the numbers on the axis
fig, subplots = setup_fig_subplots(metric)
#fig.suptitle(name)
# PDF
subplots[0].set_title('PDF')
theoretical_cdf = 0
if metric == 'Reliability':
theoretical_cdf = 1 - theoretical
sns.distplot(times,
hist=True,
ax=subplots[0],
label='Time to failures')
if metric == 'Maintainability':
theoretical_cdf = theoretical
sns.distplot(times, hist=True, ax=subplots[0], label='Time to repairs')
index = get_index_of_first_zero_in_array(1 - theoretical_cdf)
theoretical_pdf = differentiate(linspace, theoretical_cdf)
x, pdf = subplots[0].lines[0].get_data()
subplots[0].plot(x, pdf, 'b-', lw=1, alpha=0.6, label='Reconstructed')
subplots[0].plot(linspace[:index],
theoretical_pdf[:index],
'r-',
lw=1.5,
alpha=0.6,
label='Theoretical')
subplots[0].set_xlabel('time (t)')
subplots[0].set_ylabel('P(t)')
subplots[0].legend()
#subplots[0].set_xlim([0, 30])
# CDF
sns.kdeplot(times, cumulative=True, ax=subplots[1], label='Reconstructed')
subplots[1].plot(linspace[:index],
theoretical_cdf[:index],
'r-',
lw=1.5,
alpha=0.6,
label='Theoretical')
subplots[1].legend()
subplots[1].set_xlabel('time (t)')
subplots[1].set_ylabel('P(T\u2264t)')
#subplots[1].set_xlim([0, 30])
_, cdf = subplots[1].lines[0].get_data()
if metric == 'Maintainability':
subplots[1].set_title('CDF (Maintainability)')
subplots[2].plot(x,
pdf / (1 - cdf),
'b-',
lw=1.5,
alpha=0.6,
label='Reconstructed')
subplots[2].plot(linspace[:index],
theoretical_pdf[:index] /
(1 - theoretical_cdf[:index]),
'r-',
lw=1.5,
alpha=0.6,
label='Theoretical')
subplots[2].set_title('Repair Rate')
subplots[2].set_title('Repair Rate')
subplots[2].set_xlabel('time (t)')
subplots[2].set_ylabel('\u03BC(t)')
#subplots[2].set_xlim([0, 50])
subplots[2].set_ylim([0, 5])
subplots[2].legend()
#subplots[3].set_xlim([0, 320])
if metric == 'Reliability':
theoretical_reliability = theoretical
reliability = 1 - cdf
subplots[1].set_title('CDF')
#times.sort()
#samples = len(times)
#one_minus_cdf = [1 - (x / samples) for x in range(1, samples + 1)]
#cdf = [(x / samples) for x in range(1, samples + 1)]
# Get length of this and see if its the same length as the reliability of the top event
# Better yet when calculating relability function use the same number of elements as the
# length of top events time of failure or time of repair.
# Reliability
subplots[2].plot(x,
reliability,
'b-',
lw=1.5,
alpha=0.6,
label='Reconstructed')
subplots[2].plot(linspace[:index],
theoretical_reliability[:index],
'r-',
lw=1.5,
alpha=0.6,
label='Theoretical')
subplots[2].set_ylim([0, 1.05])
# For comparing graphs
subplots[2].set_title('Reliability')
subplots[2].set_xlabel('time (t)')
subplots[2].set_ylabel('R(t)')
subplots[2].legend()
subplots[3].plot(x,
pdf / reliability,
'b-',
lw=1.5,
alpha=0.6,
label='Reconstructed')
#subplots[2].set_ylim([0, 0.01])
subplots[3].plot(linspace[:index],
theoretical_pdf[:index] /
theoretical_reliability[:index],
'r-',
lw=1.5,
alpha=0.6,
label='Theoretical')
subplots[3].set_title('Failure Rate')
subplots[3].set_xlabel('time (t)')
subplots[3].set_ylabel('\u03BB(t)')
#subplots[3].set_ylim([0, 0.1])
subplots[3].legend()
#subplots[3].set_xlim([300, 325])
plt.tight_layout()
#if EXPORT_PNG is True:
fig.savefig(os.getcwd() + '/static/images/' + get_object_name(name) + '_' +
metric + '.png')
#else:
plt.show(block=False)
def plot_arbitrary__distribution_no_compare(name, metric, times):
# Reliability for now
# Maybe fix inconsistencies with the numbers on the axis
fig, subplots = setup_fig_subplots(metric)
#fig.suptitle(name + '\n\n', fontsize=16)
# PDF
subplots[0].set_title('PDF')
theoretical_cdf = 0
if metric == 'Reliability':
sns.distplot(times,
hist=True,
ax=subplots[0],
label='Time to failures')
if metric == 'Maintainability':
sns.distplot(times, hist=True, ax=subplots[0], label='Time to repairs')
x, pdf = subplots[0].lines[0].get_data()
subplots[0].plot(x, pdf, 'b-', lw=1, alpha=0.6, label='Reconstructed')
subplots[0].set_xlabel('time (t)')
subplots[0].set_ylabel('P(t)')
subplots[0].legend()
#subplots[0].set_xlim([0, 30])
# CDF
sns.kdeplot(times, cumulative=True, ax=subplots[1])
_, cdf = subplots[1].lines[0].get_data()
subplots[1].plot(x, cdf, 'b-', lw=1, alpha=0.6, label='Reconstructed')
subplots[1].set_xlabel('time (t)')
subplots[1].set_ylabel('P(T\u2264t)')
subplots[1].legend()
if metric == 'Maintainability':
subplots[1].set_title('CDF (Maintainability)')
subplots[2].plot(x,
pdf / (1 - cdf),
'b-',
lw=1.5,
alpha=0.6,
label='Reconstructed')
subplots[2].set_title('Repair Rate')
subplots[2].set_xlabel('time (t)')
subplots[2].set_ylabel('\u03BC(t)')
subplots[2].set_ylim([0, 5])
subplots[2].legend()
if metric == 'Reliability':
reliability = 1 - cdf
subplots[1].set_title('CDF')
# Get length of this and see if its the same length as the reliability of the top event
# Better yet when calculating relability function use the same number of elements as the
# length of top events time of failure or time of repair.
# Reliability
subplots[2].plot(x,
reliability,
'b-',
lw=1.5,
alpha=0.6,
label='Reconstructed')
subplots[2].set_ylim([0, 1.05])
subplots[2].legend()
# For comparing graphs
# subplots[2].set_xlim([0, 100])
subplots[2].set_title('Reliability')
subplots[2].set_xlabel('time (t)')
subplots[2].set_ylabel('R(t)')
subplots[3].plot(x,
pdf / reliability,
'b-',
lw=1.5,
alpha=0.6,
label='Reconstructed')
#subplots[2].set_ylim([0, 0.01])
subplots[3].set_title('Failure Rate')
subplots[3].set_xlabel('time (t)')
subplots[3].set_ylabel('\u03BB(t)')
#subplots[3].set_ylim([0, 0.1])
subplots[3].legend()
#subplots[3].set_xlim([300, 325])
# plt.show()
plt.tight_layout()
#if EXPORT_PNG is True:
fig.savefig(os.getcwd() + '/static/images/' + get_object_name(name) + '_' +
metric + '.png')
#else:
plt.show(block=False)
def plot_unidentified_distribution_comparison(name, metric, times,
theoretical_distribution):
# Maybe fix inconsistencies with the numbers on the axis
fig, subplots = setup_fig_subplots(metric)
linspace = calculate_linspace(theoretical_distribution)
theoretical_pdf = calculate_pdf(theoretical_distribution, linspace)
theoretical_cdf = calculate_cdf(theoretical_distribution, linspace)
theoretical_reliability = calculate_reliability(theoretical_distribution,
linspace)
if metric == 'Reliability':
sns.distplot(times,
hist=True,
ax=subplots[0],
label='Time to failures')
if metric == 'Maintainability':
sns.distplot(times, hist=True, ax=subplots[0], label='Time to repairs')
x, pdf = subplots[0].lines[0].get_data()
subplots[0].plot(x, pdf, 'b-', lw=1, alpha=0.6, label='Reconstructed')
subplots[0].set_xlabel('time (t)')
subplots[0].plot(linspace,
theoretical_pdf,
'r-',
lw=1,
alpha=0.6,
label='Theoretical')
subplots[0].set_title('PDF')
subplots[0].set_xlabel('time (t)')
subplots[0].set_ylabel('P(t)')
subplots[0].legend()
# CDF
sns.kdeplot(times, cumulative=True, ax=subplots[1])
_, cdf = subplots[1].lines[0].get_data()
# Second plot CDF and/or Maintainability
subplots[1].plot(x, cdf, 'b-', lw=1, alpha=0.6, label='Reconstructed')
subplots[1].plot(linspace,
theoretical_cdf,
'r-',
lw=1,
alpha=0.6,
label='Theoretical')
subplots[1].set_xlabel('time (t)')
subplots[1].set_ylabel('P(T\u2264t)')
subplots[1].legend()
if metric == 'Maintainability':
subplots[1].set_title('CDF (Maintainability)')
subplots[2].plot(x,
pdf / (1 - cdf),
'b-',
lw=1.5,
alpha=0.6,
label='Reconstructed')
subplots[2].plot(linspace,
theoretical_pdf / theoretical_reliability,
'r-',
lw=1,
alpha=0.6,
label='Theoretical')
subplots[2].set_title('Repair Rate')
subplots[2].set_xlabel('time (t)')
subplots[2].set_ylabel('\u03BC(t)')
subplots[2].set_ylim([0, 5])
subplots[2].legend()
if metric == 'Reliability':
reliability = 1 - cdf
subplots[1].set_title('CDF')
subplots[2].plot(x,
reliability,
'b-',
lw=1.5,
alpha=0.6,
label='Reconstructed')
subplots[2].set_ylim([0, 1.05])
# Get length of this and see if its the same length as the reliability of the top event
# Better yet when calculating relability function use the same number of elements as the
# length of top events time of failure or time of repair.
subplots[2].plot(linspace,
theoretical_reliability,
'r-',
lw=1,
alpha=0.6,
label='Theoretical')
# For comparing graphs
# subplots[2].set_xlim([0, 100])
subplots[2].set_title('Reliability')
subplots[2].set_xlabel('time (t)')
subplots[2].set_ylabel('R(t)')
subplots[2].legend()
subplots[3].plot(x,
pdf / reliability,
'b-',
lw=1.5,
alpha=0.6,
label='Reconstructed')
subplots[3].plot(linspace,
theoretical_pdf / theoretical_reliability,
'r-',
lw=1,
alpha=0.6,
label='Theoretical')
subplots[3].set_title('Failure Rate')
subplots[3].set_xlabel('time (t)')
subplots[3].set_ylabel('\u03BB(t)')
subplots[3].legend()
# plt.show()
plt.tight_layout()
#if EXPORT_PNG is True:
fig.savefig(os.getcwd() + '/static/images/' + get_object_name(name) + '_' +
metric + '.png')
plt.show(block=False)
def plot_identified_distribution_comparison(name, metric, distribution, times,
theoretical_distribution):
fig, subplots = setup_fig_subplots(metric)
#fig.suptitle(name + '\n\n', fontsize=16)
linspace = calculate_linspace(distribution)
pdf = calculate_pdf(distribution, linspace)
cdf = calculate_cdf(distribution, linspace)
theoretical_pdf = calculate_pdf(theoretical_distribution, linspace)
theoretical_cdf = calculate_cdf(theoretical_distribution, linspace)
# First plot PDF
subplots[0].plot(linspace,
pdf,
'b-',
lw=1,
alpha=0.6,
label='Reconstructed')
subplots[0].plot(linspace,
theoretical_pdf,
'r-',
lw=1,
alpha=0.6,
label='Theoretical')
subplots[0].set_xlabel('time (t)')
subplots[0].set_ylabel('P(t)')
subplots[0].legend()
if times != EMPTY_LIST:
if metric == 'Reliability':
subplots[0].hist(times,
bins=20,
normed=True,
histtype='stepfilled',
alpha=0.2,
label='Time to failures')
if metric == 'Maintainability':
subplots[0].hist(times,
bins=20,
normed=True,
histtype='stepfilled',
alpha=0.2,
label='Time to repairs')
subplots[0].legend()
subplots[0].set_title('PDF')
# Second plot CDF and/or Maintainability
subplots[1].plot(linspace,
cdf,
'b-',
lw=1,
alpha=0.6,
label='Reconstructed')
subplots[1].plot(linspace,
theoretical_cdf,
'r-',
lw=1,
alpha=0.6,
label='Theoretical')
subplots[1].set_xlabel('time (t)')
subplots[1].set_ylabel('P(T\u2264t)')
subplots[1].legend()
if metric == 'Maintainability':
subplots[1].set_title('CDF (Maintainability)')
subplots[2].plot(linspace,
pdf / (1 - cdf),
'b-',
lw=1,
alpha=0.6,
label='Reconstructed')
subplots[2].plot(linspace,
theoretical_pdf / (1 - theoretical_cdf),
'r-',
lw=1,
alpha=0.6,
label='Theoretical')
subplots[2].set_title('Repair Rate')
subplots[2].set_xlabel('time (t)')
subplots[2].set_ylabel('\u03BC(t)')
subplots[2].legend()
if distribution[0] == 'EXP':
subplots[2].set_ylim([0, 1])
# Third plot Reliability
if metric == 'Reliability':
subplots[1].set_title('CDF')
reliability = calculate_reliability(distribution, linspace)
theoretical_reliability = calculate_reliability(
theoretical_distribution, linspace)
subplots[2].plot(linspace,
reliability,
'b-',
lw=1,
alpha=0.6,
label='Reconstructed')
subplots[2].plot(linspace,
theoretical_reliability,
'r-',
lw=1,
alpha=0.6,
label='Theoretical')
subplots[2].set_xlabel('time (t)')
subplots[2].set_ylabel('R(t)')
subplots[2].legend()
subplots[2].set_title(metric)
subplots[3].plot(linspace,
pdf / reliability,
'b-',
lw=1,
alpha=0.6,
label='Reconstructed')
subplots[3].plot(linspace,
theoretical_pdf / theoretical_reliability,
'r-',
lw=1,
alpha=0.6,
label='Theoretical')
subplots[3].set_title('Failure Rate')
subplots[3].set_xlabel('time (t)')
subplots[3].set_ylabel('\u03BB(t)')
subplots[3].legend()
if distribution[0] == 'EXP':
subplots[3].set_ylim([0, 1])
# plt.show()
plt.tight_layout()
#if EXPORT_PNG is True:
fig.savefig(os.getcwd() + '/static/images/' + get_object_name(name) + '_' +
metric + '.png')
#else:
plt.show(block=False)