def report(self):
if self.valid_confidence_interval():
print "Intervalos de confiança estimados válidos!, exibindo os resultados:"
print "E[N1]: ", est.mean(self.results['m_s_N1'], self.samples)
print "IC - E[N1]", est.confidence_interval(self.results['m_s_N1'], self.results['m_s_s_N1'], self.samples)
print "E[N2]: ", est.mean(self.results['m_s_N2'], self.samples)
print "IC - E[N2]", est.confidence_interval(self.results['m_s_N2'], self.results['m_s_s_N2'], self.samples)
print "E[T1]: ", est.mean(self.results['m_s_T1'], self.samples)
print "IC - E[T1]", est.confidence_interval(self.results['m_s_T1'], self.results['m_s_s_T1'], self.samples)
print "E[T2]: ", est.mean(self.results['m_s_T2'], self.samples)
print "IC - E[T2]", est.confidence_interval(self.results['m_s_T2'], self.results['m_s_s_T2'], self.samples)
print "E[Nq1]: ", est.mean(self.results['m_s_Nq1'], self.samples)
print "IC - E[Nq1]", est.confidence_interval(self.results['m_s_Nq1'], self.results['m_s_s_Nq1'], self.samples)
print "E[Nq2]: ", est.mean(self.results['m_s_Nq2'], self.samples)
print "IC - E[Nq2]", est.confidence_interval(self.results['m_s_Nq2'], self.results['m_s_s_Nq2'], self.samples)
print "E[W1]: ", est.mean(self.results['m_s_W1'], self.samples)
print "IC - E[W1]", est.confidence_interval(self.results['m_s_W1'], self.results['m_s_s_W1'], self.samples)
print "E[W2]: ", est.mean(self.results['m_s_W2'], self.samples)
print "IC - E[W2]", est.confidence_interval(self.results['m_s_W2'], self.results['m_s_s_W2'], self.samples)
print "V(W1): ", est.mean(self.results['v_s_W1'], self.samples)
print "IC - V(W1)", est.confidence_interval(self.results['v_s_W1'], self.results['v_s_s_W1'], self.samples)
print "V(W2): ", est.mean(self.results['v_s_W2'], self.samples)
print "IC - V(W2)", est.confidence_interval(self.results['v_s_W2'], self.results['v_s_s_W2'], self.samples)
else:
print "Intervalos de confiança estimados inválidos. Aumente o numero de amostras ou de clientes por amostra"
def calc_results(self):
self.results = {
'E[N1]' : { 'value' : est.mean(self.sums['m_s_N1'], self.samples), 'c_i' : est.confidence_interval(self.sums['m_s_N1'], self.sums['m_s_s_N1'], self.samples) },
'E[N2]' : { 'value' : est.mean(self.sums['m_s_N2'], self.samples), 'c_i' : est.confidence_interval(self.sums['m_s_N2'], self.sums['m_s_s_N2'], self.samples) },
'E[T1]' : { 'value' : est.mean(self.sums['m_s_T1'], self.samples), 'c_i' : est.confidence_interval(self.sums['m_s_T1'], self.sums['m_s_s_T1'], self.samples) },
'E[T2]' : { 'value' : est.mean(self.sums['m_s_T2'], self.samples), 'c_i' : est.confidence_interval(self.sums['m_s_T2'], self.sums['m_s_s_T2'], self.samples) },
'E[Nq1]' : { 'value' : est.mean(self.sums['m_s_Nq1'], self.samples), 'c_i' : est.confidence_interval(self.sums['m_s_Nq1'], self.sums['m_s_s_Nq1'], self.samples) },
'E[Nq2]' : { 'value' : est.mean(self.sums['m_s_Nq2'], self.samples), 'c_i' : est.confidence_interval(self.sums['m_s_Nq2'], self.sums['m_s_s_Nq2'], self.samples) },
'E[W1]' : { 'value' : est.mean(self.sums['m_s_W1'], self.samples), 'c_i' : est.confidence_interval(self.sums['m_s_W1'], self.sums['m_s_s_W1'], self.samples) },
'E[W2]' : { 'value' : est.mean(self.sums['m_s_W2'], self.samples), 'c_i' : est.confidence_interval(self.sums['m_s_W2'], self.sums['m_s_s_W2'], self.samples) },
'V(W1)' : { 'value' : est.mean(self.sums['v_s_W1'], self.samples), 'c_i' : est.confidence_interval(self.sums['v_s_W1'], self.sums['v_s_s_W1'], self.samples) },
'V(W2)' : { 'value' : est.mean(self.sums['v_s_W2'], self.samples), 'c_i' : est.confidence_interval(self.sums['v_s_W2'], self.sums['v_s_s_W2'], self.samples) }
}
def process_sample(self):
s_wait_1 = 0; s_s_wait_1 = 0
s_wait_2 = 0; s_s_wait_2 = 0
s_server_1 = 0; s_server_2 = 0
# Loop que faz a soma e a soma dos quadrados dos tempos de espera e a soma dos tempos em servidor
# Dos clientes na fila 1 e na fila 2.
for client in self.clients:
s_wait_1 += client.wait(1)
s_s_wait_1 += client.wait(1)**2
s_server_1 += client.server[1]
s_wait_2 += client.wait(2)
s_s_wait_2 += client.wait(2)**2
s_server_2 += client.server[2]
# Adiciona à soma e à soma dos quadrados dos estimadores os valores estimados na rodada.
self.sums['m_s_W1'] += est.mean(s_wait_1, len(self.clients))
self.sums['m_s_s_W1'] += est.mean(s_wait_1, len(self.clients))**2
self.sums['v_s_W1'] += est.variance(s_wait_1, s_s_wait_1, len(self.clients))
self.sums['v_s_s_W1'] += est.variance(s_wait_1, s_s_wait_1, len(self.clients))**2
self.sums['m_s_N1'] += est.mean(self.N_samples['N_1'], self.t)
self.sums['m_s_s_N1'] += est.mean(self.N_samples['N_1'], self.t)**2
self.sums['m_s_Nq1'] += est.mean(self.N_samples['Nq_1'], self.t)
self.sums['m_s_s_Nq1'] += est.mean(self.N_samples['Nq_1'], self.t)**2
self.sums['m_s_T1'] += est.mean(s_wait_1, len(self.clients)) + est.mean(s_server_1, len(self.clients))
self.sums['m_s_s_T1'] += (est.mean(s_wait_1, len(self.clients)) + est.mean(s_server_1, len(self.clients)))**2
self.sums['m_s_W2'] += est.mean(s_wait_2, len(self.clients))
self.sums['m_s_s_W2'] += est.mean(s_wait_2, len(self.clients))**2
self.sums['v_s_W2'] += est.variance(s_wait_2, s_s_wait_2, len(self.clients))
self.sums['v_s_s_W2'] += est.variance(s_wait_2, s_s_wait_2, len(self.clients))**2
self.sums['m_s_N2'] += est.mean(self.N_samples['N_2'], self.t)
self.sums['m_s_s_N2'] += est.mean(self.N_samples['N_2'], self.t)**2
self.sums['m_s_Nq2'] += est.mean(self.N_samples['Nq_2'], self.t)
self.sums['m_s_s_Nq2'] += est.mean(self.N_samples['Nq_2'], self.t)**2
self.sums['m_s_T2'] += est.mean(s_wait_2, len(self.clients)) + est.mean(s_server_2, len(self.clients))
self.sums['m_s_s_T2'] += (est.mean(s_wait_2, len(self.clients)) + est.mean(s_server_2, len(self.clients)))**2
# Inicializa as estruturas de dados para a próxima rodada.
self.init_sample()
def valid_confidence_interval(self):
return (2.0*est.confidence_interval(self.results['m_s_N1'], self.results['m_s_s_N1'], self.samples) <= 0.1*est.mean(self.results['m_s_N1'], self.samples)) and \
(2.0*est.confidence_interval(self.results['m_s_N2'], self.results['m_s_s_N2'], self.samples) <= 0.1*est.mean(self.results['m_s_N2'], self.samples)) and \
(2.0*est.confidence_interval(self.results['m_s_T1'], self.results['m_s_s_T1'], self.samples) <= 0.1*est.mean(self.results['m_s_T1'], self.samples)) and \
(2.0*est.confidence_interval(self.results['m_s_T2'], self.results['m_s_s_T2'], self.samples) <= 0.1*est.mean(self.results['m_s_T2'], self.samples)) and \
(2.0*est.confidence_interval(self.results['m_s_Nq1'], self.results['m_s_s_Nq1'], self.samples) <= 0.1*est.mean(self.results['m_s_Nq1'], self.samples)) and \
(2.0*est.confidence_interval(self.results['m_s_Nq2'], self.results['m_s_s_Nq2'], self.samples) <= 0.1*est.mean(self.results['m_s_Nq2'], self.samples)) and \
(2.0*est.confidence_interval(self.results['m_s_W1'], self.results['m_s_s_W1'], self.samples) <= 0.1*est.mean(self.results['m_s_W1'], self.samples)) and \
(2.0*est.confidence_interval(self.results['m_s_W2'], self.results['m_s_s_W2'], self.samples) <= 0.1*est.mean(self.results['m_s_W2'], self.samples)) and \
(2.0*est.confidence_interval(self.results['v_s_W1'], self.results['v_s_s_W1'], self.samples) <= 0.1*est.mean(self.results['v_s_W1'], self.samples)) and \
(2.0*est.confidence_interval(self.results['v_s_W2'], self.results['v_s_s_W2'], self.samples) <= 0.1*est.mean(self.results['v_s_W2'], self.samples))
def process_sample():
s_wait_1 = 0; s_s_wait_1 = 0
s_wait_2 = 0; s_s_wait_2 = 0
s_server_1 = 0; s_server_2 = 0
for client in self.clients:
s_wait_1 += client.wait(1)
s_s_wait_1 += client.wait(1)**2
s_server_1 += client.server[1]
s_wait_2 += client.wait(2)
s_s_wait_2 += client.wait(2)**2
s_server_2 += client.server[2]
self.results['m_s_W1'] += est.mean(s_wait_1, len(self.clients))
self.results['m_s_s_W1'] += est.mean(s_wait_1, len(self.clients))**2
self.results['v_s_W1'] += est.variance(s_wait_1, s_s_wait_1, len(self.clients))
self.results['v_s_s_W1'] += est.variance(s_wait_1, s_s_wait_1, len(self.clients))**2
self.results['m_s_N1'] += est.mean(self.N_samples['N_1'], self.t)
self.results['m_s_s_N1'] += est.mean(self.N_samples['N_1'], self.t)**2
self.results['m_s_Nq1'] += est.mean(self.N_samples['Nq_1'], self.t)
self.results['m_s_s_Nq1'] += est.mean(self.N_samples['Nq_1'], self.t)**2
self.results['m_s_T1'] += est.mean(s_wait_1, len(self.clients)) + est.mean(s_server_1, len(self.clients))
self.results['m_s_s_T1'] += (est.mean(s_wait_1, len(self.clients)) + est.mean(s_server_1, len(self.clients)))**2
self.results['m_s_W2'] += est.mean(s_wait_2, len(self.clients))
self.results['m_s_s_W2'] += est.mean(s_wait_2, len(self.clients))**2
self.results['v_s_W2'] += est.variance(s_wait_2, s_s_wait_2, len(self.clients))
self.results['v_s_s_W2'] += est.variance(s_wait_2, s_s_wait_2, len(self.clients))**2
self.results['m_s_N2'] += est.mean(self.N_samples['N_2'], self.t)
self.results['m_s_s_N2'] += est.mean(self.N_samples['N_2'], self.t)**2
self.results['m_s_Nq2'] += est.mean(self.N_samples['Nq_2'], self.t)
self.results['m_s_s_Nq2'] += est.mean(self.N_samples['Nq_2'], self.t)**2
self.results['m_s_T2'] += est.mean(s_wait_2, len(self.clients)) + est.mean(s_server_2, len(self.clients))
self.results['m_s_s_T2'] += (est.mean(s_wait_2, len(self.clients)) + est.mean(s_server_2, len(self.clients)))**2
self.init_sample()
