def update(self, time, current_state, next_state):
""" updates the discounted total cost and health utility
:param time: simulation time
:param current_state: current health state
:param next_state: next health state
"""
# cost and utility (per unit of time) during the period since the last recording until now
cost = self.params.annualStateCosts[
current_state.value] + self.params.annualTreatmentCost
utility = self.params.annualStateUtilities[current_state.value]
# discounted cost and utility (continuously compounded)
discounted_cost = Econ.pv_continuous_payment(
payment=cost,
discount_rate=self.params.discountRate,
discount_period=(self.tLastRecorded, time))
discounted_utility = Econ.pv_continuous_payment(
payment=utility,
discount_rate=self.params.discountRate,
discount_period=(self.tLastRecorded, time))
# update total discounted cost and utility
self.totalDiscountedCost += discounted_cost
self.totalDiscountedUtility += discounted_utility
# update the time since last recording to the current time
self.tLastRecorded = time
def report_CEA_CBA(multi_cohort_outcomes_SOC, multi_cohort_outcomes_NSB):
""" performs cost-effectiveness and cost-benefit analyses
:param multi_cohort_outcomes_SOC: outcomes of a multi-cohort simulated under SOC diagnostic
:param multi_cohort_outcomes_NSB: outcomes of a multi-cohort simulated under NSB diagnostic
"""
# define two strategies
SOC_diagnostic_strategy = Econ.Strategy(
name='SOC Diagnostic',
cost_obs=multi_cohort_outcomes_SOC.meanCosts,
effect_obs=multi_cohort_outcomes_SOC.totalYLL,
color='green'
)
NSB_diagnostic_strategy = Econ.Strategy(
name='NSB Diagnostic',
cost_obs=multi_cohort_outcomes_NSB.meanCosts,
effect_obs=multi_cohort_outcomes_NSB.totalYLL,
color='blue'
)
# do CEA
CEA = Econ.CEA(
strategies=[SOC_diagnostic_strategy, NSB_diagnostic_strategy],
if_paired=True,
health_measure='d'
)
# show the cost-effectiveness plane
show_ce_figure(CEA=CEA)
# report the CE table
CEA.build_CE_table(
interval_type='p', # prediction intervals
alpha=D.ALPHA,
cost_digits=0,
effect_digits=2,
icer_digits=2)
# CBA
NBA = Econ.CBA(
strategies=[SOC_diagnostic_strategy, NSB_diagnostic_strategy],
if_paired=True,
health_measure='d'
)
# show the net monetary benefit figure
NBA.graph_incremental_NMBs(
min_wtp=0,
max_wtp=50000,
title='Cost-Benefit Analysis',
x_label='Willingness-To-Pay for One Additional QALY ($)',
y_label='Incremental Net Monetary Benefit ($)',
interval_type='p',
show_legend=True,
figure_size=(6, 5)
)
def report_CEA_CBA(multi_trees_P, multi_trees_N):
""" performs cost-effectiveness and cost-benefit analyses
:param multi_cohort_outcomes_mono: outcomes of a multi-cohort simulated under mono therapy
:param multi_cohort_outcomes_combo: outcomes of a multi-cohort simulated under combination therapy
"""
# define two strategies
Palivizumab_therapy_strategy = Econ.Strategy(
name='Palivizumab Therapy',
cost_obs=multi_trees_P.PCosts,
effect_obs=multi_trees_P.PQALYs,
color='green'
)
No_prophylaxis_strategy = Econ.Strategy(
name='No Prophylaxis',
cost_obs=multi_trees_N.NCosts,
effect_obs=multi_trees_N.NQALYs,
color='blue'
)
# do CEA
CEA = Econ.CEA(
strategies=[Palivizumab_therapy_strategy, No_prophylaxis_strategy],
if_paired=True
)
# show the cost-effectiveness plane
show_ce_figure(CEA=CEA)
# report the CE table
CEA.build_CE_table(
interval_type='p', # prediction intervals
alpha=D.ALPHA,
cost_digits=0,
effect_digits=2,
icer_digits=2)
# CBA
NBA = Econ.CBA(
strategies=[Palivizumab_therapy_strategy, No_prophylaxis_strategy],
if_paired=True
)
# show the net monetary benefit figure
NBA.graph_incremental_NMBs(
min_wtp=0,
max_wtp=5000000,
title='Cost-Benefit Analysis',
x_label='Willingness-To-Pay for One Additional QALY ($)',
y_label='Incremental Net Monetary Benefit ($)',
interval_type='p',
show_legend=True,
figure_size=(6, 5)
)
def report_CEA_CBA(multi_cohort_outcomes_no, multi_cohort_outcomes_asp):
# define two strategies
no_therapy_strategy = Econ.Strategy(
name='Mono Therapy',
cost_obs=multi_cohort_outcomes_no.meanCosts,
effect_obs=multi_cohort_outcomes_asp.meanQALYs,
color='green'
)
asp_therapy_strategy = Econ.Strategy(
name='Combination Therapy',
cost_obs=multi_cohort_outcomes_asp.meanCosts,
effect_obs=multi_cohort_outcomes_no.meanQALYs,
color='blue'
)
# do CEA
CEA = Econ.CEA(
strategies=[no_therapy_strategy, asp_therapy_strategy],
if_paired=True
)
# show the cost-effectiveness plane
show_ce_figure(CEA=CEA)
# report the CE table
CEA.build_CE_table(
interval_type='p', # prediction intervals
alpha=D.ALPHA,
cost_digits=0,
effect_digits=2,
icer_digits=2)
# CBA
NBA = Econ.CBA(
strategies=[no_therapy_strategy, asp_therapy_strategy],
if_paired=True
)
# show the net monetary benefit figure
NBA.graph_incremental_NMBs(
min_wtp=0,
max_wtp=50000,
title='Cost-Benefit Analysis',
x_label='Willingness-To-Pay for One Additional QALY ($)',
y_label='Incremental Net Monetary Benefit ($)',
interval_type='p',
show_legend=True,
figure_size=(6, 5)
)
def update(self, k, current_state, next_state):
""" updates the discounted total cost and health utility
:param k: simulation time step
:param current_state: current health state
:param next_state: next health state
"""
# update cost
cost = 0.5 * (self.params.annualStateCosts[current_state.value] +
self.params.annualStateCosts[next_state.value])
# update utility
# utility = 0.5 * (self.params.annualStateUtilities[current_state.value] +
# self.params.annualStateUtilities[next_state.value])
# add the cost of treatment
# if Chron's death will occur, add the cost for half-year of treatment
if next_state == P.HealthStates.DEATH:
cost += 0.5 * self.params.annualTreatmentCost
else:
cost += 1 * self.params.annualTreatmentCost
# update total discounted cost and utility (corrected for the half-cycle effect)
self.totalDiscountedCost += Econ.pv_single_payment(
payment=cost,
discount_rate=self.params.discountRate / 2,
discount_period=2 * k + 1)
def update(self, k, current_state, next_state):
""" updates the discounted total cost and health utility
:param k: simulation time step
:param current_state: current health state
:param next_state: next health state
"""
# update cost
cost = (self.params.annualTotalCosts[current_state.value] +
self.params.annualTotalCosts[next_state.value])
# # update utility
# utility = 0.5 * (self.params.annualTotalUtility[current_state.value] +
# self.params.annualTotalUtility[next_state.value])
# add the cost of treatment
if current_state == P.HealthStates.ACTIVE_TB and next_state == P.HealthStates.ACTIVE_TB:
cost += 1 * self.params.annualTreatmentCost
elif current_state == P.HealthStates.ACTIVE_TB and next_state == P.HealthStates.CURED or P.HealthStates.INCOMPLETE:
cost += 0.5 * self.params.annualTreatmentCost
# update total discounted cost and utility (corrected for the half-cycle effect)
self.totalDiscountedCost += Econ.pv_single_payment(
payment=cost,
discount_rate=self.params.discountRate,
discount_period=k + 1)
def report_CEA_CBA(sim_outcomes_amino, sim_outcomes_immuno):
""" performs cost-effectiveness and cost-benefit analyses
:param sim_outcomes_amino: outcomes of a cohort simulated under aminosalicylate therapy
:param sim_outcomes_immuno: outcomes of a cohort simulated under immunosuppresive therapy
"""
# define two strategies
amino_therapy_strategy = Econ.Strategy(
name='Aminosalicylate Therapy',
cost_obs=sim_outcomes_amino.costs,
effect_obs=sim_outcomes_amino.numPatientsAlive,
color='green')
immuno_therapy_strategy = Econ.Strategy(
name='Immunosuppresive Therapy',
cost_obs=sim_outcomes_immuno.costs,
effect_obs=sim_outcomes_immuno.numPatientsAlive,
color='blue')
# do CEA
CEA = Econ.CEA(
strategies=[amino_therapy_strategy, immuno_therapy_strategy],
if_paired=False)
# show the cost-effectiveness plane
show_ce_figure(CEA=CEA)
# report the CE table
CEA.build_CE_table(interval_type='c',
alpha=D.ALPHA,
cost_digits=0,
effect_digits=2,
icer_digits=2)
# CBA
NBA = Econ.CBA(
strategies=[amino_therapy_strategy, immuno_therapy_strategy],
if_paired=False)
# show the net monetary benefit figure
NBA.graph_incremental_NMBs(
min_wtp=0,
max_wtp=50000,
title='Cost-Benefit Analysis',
x_label='Willingness-to-pay for one additional QALY ($)',
y_label='Incremental Net Monetary Benefit ($)',
interval_type='c',
show_legend=True,
figure_size=(6, 5))
def report_CEA_CBA(sim_outcomes_SOC, sim_outcomes_NSB):
""" performs cost-effectiveness and cost-benefit analyses
:param sim_outcomes_mono: outcomes of a cohort simulated under mono therapy
:param sim_outcomes_combo: outcomes of a cohort simulated under combination therapy
"""
# define two strategies
SOC_diagnostic_strategy = Econ.Strategy(
name='SOC Diagnostic',
cost_obs=sim_outcomes_SOC.costsPresenting,
effect_obs=sim_outcomes_SOC.listYLLPresenting,
color='green')
sim_outcomes_NSB = Econ.Strategy(
name='NSB Diagnostic',
cost_obs=sim_outcomes_NSB.costsPresenting,
effect_obs=sim_outcomes_NSB.listYLLPresenting,
color='blue')
# do CEA
CEA = Econ.CEA(strategies=[SOC_diagnostic_strategy, sim_outcomes_NSB],
if_paired=False,
health_measure='d')
# show the cost-effectiveness plane
show_ce_figure(CEA=CEA)
# report the CE table
CEA.build_CE_table(interval_type='c',
alpha=D.ALPHA,
cost_digits=0,
effect_digits=2,
icer_digits=2)
# CBA
NBA = Econ.CBA(strategies=[SOC_diagnostic_strategy, sim_outcomes_NSB],
if_paired=False)
# show the net monetary benefit figure
NBA.graph_incremental_NMBs(
min_wtp=0,
max_wtp=50000,
title='Cost-Benefit Analysis',
x_label='Willingness-to-pay for one additional YLL ($)',
y_label='Incremental Net Monetary Benefit ($)',
interval_type='c',
show_legend=True,
figure_size=(6, 5))
def report_CEA_CBA(sim_outcomes_none, sim_outcomes_treat):
""" performs cost-effectiveness and cost-benefit analyses
:param sim_outcomes_mono: outcomes of a cohort simulated under mono therapy
:param sim_outcomes_combo: outcomes of a cohort simulated under combination therapy
"""
# define two strategies
none_therapy_strategy = Econ.Strategy(
name='No Therapy',
cost_obs=sim_outcomes_none.costs,
effect_obs=sim_outcomes_none.nTotalCured,
color='green')
treat_therapy_strategy = Econ.Strategy(
name='Intervention to Increase Adherence',
cost_obs=sim_outcomes_treat.costs,
effect_obs=sim_outcomes_treat.nTotalCured,
color='blue')
# do CEA
CEA = Econ.CEA(strategies=[none_therapy_strategy, treat_therapy_strategy],
if_paired=False)
# show the cost-effectiveness plane
show_ce_figure(CEA=CEA)
# report the CE table
CEA.build_CE_table(interval_type='c',
alpha=D.ALPHA,
cost_digits=0,
effect_digits=2,
icer_digits=2)
# CBA
NBA = Econ.CBA(strategies=[none_therapy_strategy, treat_therapy_strategy],
if_paired=False)
# show the net monetary benefit figure
NBA.graph_incremental_NMBs(
min_wtp=0,
max_wtp=1000,
title='Cost-Benefit Analysis',
x_label='Willingness-to-pay for one additional patient cured ($)',
y_label='Incremental Net Monetary Benefit ($)',
interval_type='c',
show_legend=True,
figure_size=(6, 5))
def report_CEA_CBA(sim_outcomes_NO, sim_outcomes_ASP):
# define two strategies
NO_therapy_strategy = Econ.Strategy(name='No Therapy',
cost_obs=sim_outcomes_NO.costs,
effect_obs=sim_outcomes_NO.utilities,
color='green')
ASP_therapy_strategy = Econ.Strategy(name='Aspirin Therapy',
cost_obs=sim_outcomes_ASP.costs,
effect_obs=sim_outcomes_ASP.utilities,
color='blue')
# do CEA
CEA = Econ.CEA(strategies=[NO_therapy_strategy, ASP_therapy_strategy],
if_paired=False)
# show the cost-effectiveness plane
show_ce_figure(CEA=CEA)
# report the CE table
CEA.build_CE_table(interval_type='c',
alpha=D.ALPHA,
cost_digits=0,
effect_digits=2,
icer_digits=2)
# CBA
NBA = Econ.CBA(strategies=[NO_therapy_strategy, ASP_therapy_strategy],
if_paired=False)
# show the net monetary benefit figure
NBA.graph_incremental_NMBs(
min_wtp=0,
max_wtp=100000,
title='Cost-Benefit Analysis',
x_label='Willingness-to-pay for one additional QALY ($)',
y_label='Incremental Net Monetary Benefit ($)',
interval_type='c',
show_legend=True,
figure_size=(6, 5))
from SimPy import EconEvalClasses as ce
import numpy as np
np.random.seed(573)
s_center = np.random.normal(0, 5000, (10, 2))
s0 = ce.Strategy("s1", s_center[0, 0] + np.random.normal(0, 200, 10),
s_center[0, 1] + np.random.normal(0, 200, 10))
s1 = ce.Strategy("s2", s_center[1, 0] + np.random.normal(0, 200, 10),
s_center[1, 1] + np.random.normal(0, 100, 10))
s2 = ce.Strategy("s3", s_center[2, 0] + np.random.normal(0, 200, 10),
s_center[2, 1] + np.random.normal(0, 200, 10))
s3 = ce.Strategy("s4", s_center[3, 0] + np.random.normal(0, 200, 10),
s_center[3, 1] + np.random.normal(0, 200, 10))
s4 = ce.Strategy("s5", s_center[4, 0] + np.random.normal(0, 200, 10),
s_center[4, 1] + np.random.normal(0, 200, 10))
s5 = ce.Strategy("s6", s_center[5, 0] + np.random.normal(0, 200, 10),
s_center[5, 1] + np.random.normal(0, 200, 10))
s6 = ce.Strategy("s7", s_center[6, 0] + np.random.normal(0, 200, 10),
s_center[6, 1] + np.random.normal(0, 200, 10))
s7 = ce.Strategy("s8", s_center[7, 0] + np.random.normal(0, 200, 10),
s_center[7, 1] + np.random.normal(0, 200, 10))
s8 = ce.Strategy("s9", s_center[8, 0] + np.random.normal(0, 200, 10),
s_center[8, 1] + np.random.normal(0, 200, 10))
s9 = ce.Strategy("s10", s_center[9, 0] + np.random.normal(0, 200, 10),
s_center[9, 1] + np.random.normal(0, 200, 10))
# create a CEA object -- unpaired
myCEA = ce.CEA([s0, s1, s2, s3, s4, s5, s6, s7, s8, s9], if_paired=False)
# plot with label and sample cloud
from SimPy import EconEvalClasses as EV
S0 = EV.Strategy(name='Base', cost_obs=[100], effect_obs=[1])
S1 = EV.Strategy(name='A1', cost_obs=[800], effect_obs=[0.5])
S2 = EV.Strategy(name='A2', cost_obs=[2000], effect_obs=[10])
S3 = EV.Strategy(name='A3', cost_obs=[500], effect_obs=[7])
S4 = EV.Strategy(name='A4', cost_obs=[-100], effect_obs=[2])
cea = EV.CEA(strategies=[S0, S1, S2, S3, S4],
if_paired=False,
health_measure=EV.HealthMeasure.UTILITY)
print('On frontier')
for s in cea.get_strategies_on_frontier():
print(s.name)
print('Not on frontier')
for s in cea.get_strategies_not_on_frontier():
print(s.name)
cea.show_CE_plane('CE plane',
'E[Effect]',
'E[Cost]',
show_names=True,
figure_size=6)
cea.build_CE_table(cost_digits=0, interval_type='n')
alpha=0.05
# summary statistics for all conditions
print('Summary Stats OS Cost', simulation.get_sumStat_OS_cost().get_mean())
print('95% Confidence Interval for OS Cost', simulation.get_sumStat_OS_cost().get_t_CI(alpha))
print('Summary Stats No OS Cost', simulation.get_sumStat_NoOS_cost().get_mean())
print('95% Confidence Interval for No OS Cost', simulation.get_sumStat_NoOS_cost().get_t_CI(alpha))
print('Summary Stats OS Utility', simulation.get_sumStat_OS_utility().get_mean())
print('95% Confidence Interval for OS Utility', simulation.get_sumStat_OS_utility().get_t_CI(alpha))
print('Summary Stats No OS Utility', simulation.get_sumStat_NoOS_utility().get_mean())
print('95% Confidence Interval for No OS Utility', simulation.get_sumStat_NoOS_utility().get_t_CI(alpha))
print(" ")
print (" ")
# CEA plot
# currently turning out weird but hopefully will have more of a 'cloud' when we randomize the simulation parameters
s1 = ce.Strategy('Intervention', cost_obs=simulation.get_OS_costs(), effect_obs=simulation.get_OS_utilities())
s2 = ce.Strategy('No Intervention', cost_obs=simulation.get_NoOS_costs(), effect_obs=simulation.get_NoOS_utilities())
myCEA = ce.CEA([s2, s1], if_paired=False)
myCEA.show_CE_plane('CE Plane with cost vs utilities, Bonanza', x_label='Utilities', y_label='Costs',
show_legend=True, show_clouds=True, figure_size=6)
# ICER table
print('')
# return none and write result into csv
print(myCEA.build_CE_table(ce.Interval.PREDICTION))
print(myCEA.build_CE_table())
import SimPy.EconEvalClasses as EconEval
import numpy as np
np.random.seed(573)
cost_base = np.random.normal(loc=10000, scale=100, size=1000)
effect_base = np.random.normal(loc=2, scale=.1, size=1000)
cost_intervention = np.random.normal(loc=20000, scale=200, size=1000)
effect_intervention = np.random.normal(loc=1, scale=.2, size=1000)
print('')
# ICER calculation assuming paired observations
ICER_paired = EconEval.ICER_paired('Testing paired ICER', cost_intervention,
effect_intervention, cost_base, effect_base,
EconEval.HealthMeasure.DISUTILITY)
print('Paired ICER (confidence and prediction interval): ',
ICER_paired.get_ICER(), ICER_paired.get_CI(0.05, 1000),
ICER_paired.get_PI(0.05, ))
# ICER calculation assuming independent observations
ICER_indp = EconEval.ICER_indp('Testing independent ICER', cost_intervention,
effect_intervention, cost_base, effect_base,
EconEval.HealthMeasure.DISUTILITY)
print('Independent ICER (confidence and prediction interval): ',
ICER_indp.get_ICER(), ICER_indp.get_CI(0.05, 1000),
ICER_indp.get_PI(0.05, ))
# try NMB
NMB_paired = EconEval.NMB_paired("Testing paired NMB", cost_intervention,
effect_intervention, cost_base, effect_base,
import SimPy.EconEvalClasses as EconEval
import numpy as np
np.random.seed(573)
cost_base = np.random.normal(loc=10000, scale=100, size=1000)
effect_base = np.random.normal(loc=1, scale=.1, size=1000)
cost_intervention = np.random.normal(loc=20000, scale=200, size=1000)
effect_intervention = np.random.normal(loc=2, scale=.2, size=1000)
print('')
# ICER calculation assuming paired observations
ICER_paired = EconEval.ICER_paired('Testing paired ICER',
cost_intervention, effect_intervention, cost_base, effect_base)
print('Paired ICER (confidence and prediction interval): ',
ICER_paired.get_ICER(),
ICER_paired.get_CI(0.05, 1000),
ICER_paired.get_PI(0.05, ))
# ICER calculation assuming independent observations
ICER_indp = EconEval.ICER_indp('Testing independent ICER',
cost_intervention, effect_intervention, cost_base, effect_base)
print('Independent ICER (confidence and prediction interval): ',
ICER_indp.get_ICER(),
ICER_indp.get_CI(0.05, 1000),
ICER_indp.get_PI(0.05, ))
# try NMB
NMB_paired = EconEval.NMB_paired("Testing paired NMB",
cost_intervention, effect_intervention, cost_base, effect_base)
from SimPy import EconEvalClasses as EV
import numpy as np
np.random.seed(seed=1)
S0 = EV.Strategy(name='Base',
cost_obs=np.random.normal(loc=100, scale=10, size=100),
effect_obs=np.random.normal(loc=1, scale=.2, size=100))
S1 = EV.Strategy(name='A1',
cost_obs=np.random.normal(loc=800, scale=50, size=100),
effect_obs=np.random.normal(loc=0.5, scale=0.1, size=100))
S2 = EV.Strategy(name='A2',
cost_obs=np.random.normal(loc=2000, scale=200, size=100),
effect_obs=np.random.normal(loc=10, scale=1, size=100))
S3 = EV.Strategy(name='A3',
cost_obs=np.random.normal(loc=500, scale=50, size=100),
effect_obs=np.random.normal(loc=7, scale=1, size=100))
S4 = EV.Strategy(name='A4',
cost_obs=np.random.normal(loc=-100, scale=10, size=100),
effect_obs=np.random.normal(loc=2, scale=0.1, size=100))
cea = EV.CEA(strategies=[S0, S1, S2, S3, S4], if_paired=False, health_measure='u')
print('On frontier')
for s in cea.get_strategies_on_frontier():
print(s.name)
print('Not on frontier')
for s in cea.get_strategies_not_on_frontier():
print(s.name)
import SimPy.EconEvalClasses as Econ
print('Present value of $10 collected in year 20 at discount rate 5%:')
print(
Econ.pv_single_payment(payment=10, discount_rate=0.05, discount_period=20))
print(
'\nPresent value of $10 collected in year 20 at discount rate 5% (discounted continuously):'
)
print('These 2 numbers should be almost the same:')
print(
Econ.pv_single_payment(payment=10,
discount_rate=0.05,
discount_period=20,
discount_continuously=True))
print(
Econ.pv_single_payment(payment=10,
discount_rate=0.05 / 100,
discount_period=20 * 100,
discount_continuously=False))
print(
'\nPresent value of a continuous payment of $10 over the period [10, 20] at discount rate 5%:'
)
print(
Econ.pv_continuous_payment(payment=10,
discount_rate=0.05,
discount_period=(10, 20)))
print(
'\nPresent value of a continuous payment of $10 over the period [10, 20] at discount rate 0%:'
)
print('Summary Stats OS Utility',
simulation.get_sumStat_OS_utility().get_mean())
print('95% Confidence Interval for OS Utility',
simulation.get_sumStat_OS_utility().get_t_CI(alpha))
print('Summary Stats No OS Utility',
simulation.get_sumStat_NoOS_utility().get_mean())
print('95% Confidence Interval for No OS Utility',
simulation.get_sumStat_NoOS_utility().get_t_CI(alpha))
print(" ")
print(" ")
# CEA plot
# currently turning out weird but hopefully will have more of a 'cloud' when we randomize the simulation parameters
s1 = ce.Strategy('OpSmile',
cost_obs=simulation.get_OS_costs(),
effect_obs=simulation.get_OS_utilities())
s2 = ce.Strategy('No OpSmile',
cost_obs=simulation.get_NoOS_costs(),
effect_obs=simulation.get_NoOS_utilities())
myCEA = ce.CEA([s1, s2],
if_paired=False) # double check to see if this is paired or not
myCEA.show_CE_plane('CE Plane with cost vs utilities',
x_label='Cost',
y_label='Utilities',
show_legend=True,
show_clouds=True,
figure_size=6)
#look at class material to see how simulation parameters might have been used to form a cloud
from SimPy import EconEvalClasses as ce
import numpy as np
np.random.seed(573)
s_center = np.array([[10000, 0.2], [20000, 0.3], [50000, 0.35]])
s0 = ce.Strategy("s1", s_center[0, 0] + np.random.normal(0, 1000, 10),
s_center[0, 1] + np.random.normal(0, 0.01, 10))
s1 = ce.Strategy("s1", s_center[1, 0] + np.random.normal(0, 1000, 10),
s_center[1, 1] + np.random.normal(0, 0.01, 10))
s2 = ce.Strategy("s2", s_center[2, 0] + np.random.normal(0, 1000, 10),
s_center[2, 1] + np.random.normal(0, 0.05, 10))
nmb_paired = ce.CBA([s0, s1, s2],
if_paired=True) # list of frontier strategies as input
nmb_indp = ce.CBA([s0, s1, s2],
if_paired=False) # list of frontier strategies as input
# Try NMB_Lines figure - paired CI
nmb_paired.graph_deltaNMB_lines(0,
50000,
"deltaNMB lines for paired CI",
"wtp values",
"NMB values",
interval_type=ce.Interval.CONFIDENCE,
show_legend=True,
figure_size=6)
# Try NMB_Lines figure - paired PI
nmb_paired.graph_deltaNMB_lines(0,
50000,