def get_aggregate_demand(self, p1, p2):
""" Aggregate demand of the individual agents to the overall demand
for each good in the economy.
"""
# Antibugging
integrity_checks('get_aggregate_demand_in', p1, p2)
# Distribute class attributes
agent_objs = self.population
# Initialize result container
rslt = np.array([0.0, 0.0])
# Loop over all agents in the population.
for agent_obj in agent_objs:
agent_obj.choose(p1, p2)
demand = agent_obj.get_individual_demand()
rslt += demand
# Type conversion
rslt = rslt.tolist()
# Quality Assurance
integrity_checks('get_aggregate_demand_out', rslt)
# Finishing
return rslt
def get_aggregate_demand(self, p1, p2):
""" Aggregate demand of the individual agents to the overall demand
for each good in the economy.
"""
# Antibugging
integrity_checks('get_aggregate_demand_in', p1, p2)
# Distribute class attributes
agent_objs = self.population
# Loop over all agents in the population.
demand_x1 = []
demand_x2 = []
for agent_obj in agent_objs:
agent_obj.choose(p1, p2)
demand_x1.append(agent_obj.get_individual_demand()[0])
demand_x2.append(agent_obj.get_individual_demand()[1])
# Record output
rslt = {
'demand': [np.sum(demand_x1), np.sum(demand_x2)],
'sd': [np.std(demand_x1), np.std(demand_x2)]
}
# Quality Assurance
integrity_checks('get_aggregate_demand_out', rslt)
# Finishing
return rslt
def get_aggregate_demand(self, p1, p2):
""" Aggregate demand of the individual agents to the overall demand
for each good in the economy.
"""
# Antibugging
integrity_checks('get_aggregate_demand_in', p1, p2)
# Distribute class attributes
agent_objs = self.population
# Loop over all agents in the population.
demand_x1 = []
demand_x2 = []
for agent_obj in agent_objs:
agent_obj.choose(p1, p2)
demand_x1.append(agent_obj.get_individual_demand()[0])
demand_x2.append(agent_obj.get_individual_demand()[1])
# Record output
rslt = {'demand': [np.sum(demand_x1), np.sum(demand_x2)], 'sd': [np.std(demand_x1), np.std(demand_x2)]}
# Quality Assurance
integrity_checks('get_aggregate_demand_out', rslt)
# Finishing
return rslt
def set_preference_parameter(self, alpha):
""" Set the preference parameter.
"""
# Antibugging
integrity_checks('set_preference_parameter', alpha)
# Attach preference parameter as class attribute
self.alpha = alpha
def set_endowment(self, y):
""" Set the endowment.
"""
# Antibugging
integrity_checks('set_endowment', y)
# Attach endowment as class attribute
self.y = y
def set_preference_parameter(self, alpha):
""" Set the preference parameter.
"""
# Antibugging
integrity_checks('set_preference_parameter', alpha)
# Attach preference parameter as class attribute
self.alpha = alpha
def set_endowment(self, y):
""" Set the endowment.
"""
# Antibugging
integrity_checks('set_endowment', y)
# Attach endowment as class attribute
self.y = y
def __init__(self, agent_objs):
""" Initialize instance of economy class and attach the agent
population as a class attribute.
"""
# Antibugging
integrity_checks('__init__', agent_objs)
# Attach initialization attributes
self.population = agent_objs
def __init__(self, agent_objs):
""" Initialize instance of economy class and attach the agent
population as a class attribute.
"""
# Antibugging
integrity_checks('__init__', agent_objs)
# Attach initialization attributes
self.population = agent_objs
def get_individual_demand(self):
""" Get the agents demand for the goods.
"""
# Extract demand from class attributes
rslt = self.x[:]
# Quality Checks
integrity_checks('get_individual_demand', rslt)
# Finishing
return rslt
def get_individual_demand(self):
""" Get the agents demand for the goods.
"""
# Extract demand from class attributes
rslt = self.x[:]
# Quality Checks
integrity_checks('get_individual_demand', rslt)
# Finishing
return rslt
def spending(x, p1, p2):
""" Calculate spending level.
"""
# Antibugging
integrity_checks('spending', x, p1, p2)
# Distribute demands
x1, x2 = x
# Calculate expenses
e = x1 * p1 + x2 * p2
# Finishing
return e
def _criterion(self, x):
""" Evaluate criterion function.
"""
# Antibugging
integrity_checks('_criterion', x)
# Ensure non-negativity of demand
x = x**2
# Utility calculation
u = self.get_utility(x)
# Finishing
return -u
def _criterion(self, x):
""" Evaluate criterion function.
"""
# Antibugging
integrity_checks('_criterion', x)
# Ensure non-negativity of demand
x = x ** 2
# Utility calculation
u = self.get_utility(x)
# Finishing
return -u
def choose(self, p1, p2):
""" Choose the desired bundle of goods for different agent
decision rules.
"""
# Antibugging
integrity_checks('choose', p1, p2)
# Distribute class attributes
y = self.y
x = self._choose(y, p1, p2)
# Update class attributes
self.x = x
def choose(self, p1, p2):
""" Choose the desired bundle of goods for different agent
decision rules.
"""
# Antibugging
integrity_checks('choose', p1, p2)
# Distribute class attributes
y = self.y
x = self._choose(y, p1, p2)
# Update class attributes
self.x = x
def spending(x, p1, p2):
""" Calculate spending level.
"""
# Antibugging
integrity_checks('spending', x, p1, p2)
# Distribute demands
x1, x2 = x
# Calculate expenses
e = x1 * p1 + x2 * p2
# Finishing
return e
def _constraint(self, x, p1, p2):
""" Non-negativity constraint for the SLSQP algorithm.
"""
# Antibugging
integrity_checks('_constraint', x, p1, p2)
# Distribute endowment
y = self.y
# Ensure non-negativity
x = x ** 2
# Calculate savings
cons = y - self.spending(x, p1, p2)
# Finishing
return cons
def _constraint(self, x, p1, p2):
""" Non-negativity constraint for the SLSQP algorithm.
"""
# Antibugging
integrity_checks('_constraint', x, p1, p2)
# Distribute endowment
y = self.y
# Ensure non-negativity
x = x**2
# Calculate savings
cons = y - self.spending(x, p1, p2)
# Finishing
return cons
def _choose(self, y, p1, p2):
""" Choose utility-maximizing bundle.
"""
# Antibugging
integrity_checks('_choose_rational_in', y, p1, p2)
# Determine starting values
x0 = np.array([(0.5 * y) / p1, (0.5 * y) / p2])
# Construct budget constraint
constraint_divergence = dict()
constraint_divergence['type'] = 'eq'
constraint_divergence['args'] = (p1, p2)
constraint_divergence['fun'] = self._constraint
constraints = [
constraint_divergence,
]
# Call constraint-optimizer. Of course, we could determine the
# optimal bundle directly, but I wanted to illustrate the use of
# a constraint optimization algorithm to you.
rslt = minimize(self._criterion,
x0,
method='SLSQP',
constraints=constraints)
# Check for convergence
assert (rslt['success'] == True)
# Transformation of result.
x = rslt['x']**2
# Type conversion
x = x.tolist()
# Quality Checks
integrity_checks('_choose_rational_out', x)
# Finishing
return x
def _choose(self, y, p1, p2):
""" Choose utility-maximizing bundle.
"""
# Antibugging
integrity_checks('_choose_rational_in', y, p1, p2)
# Determine starting values
x0 = np.array([(0.5 * y) / p1, (0.5 * y) / p2])
# Construct budget constraint
constraint_divergence = dict()
constraint_divergence['type'] = 'eq'
constraint_divergence['args'] = (p1, p2)
constraint_divergence['fun'] = self._constraint
constraints = [constraint_divergence, ]
# Call constraint-optimizer. Of course, we could determine the
# optimal bundle directly, but I wanted to illustrate the use of
# a constraint optimization algorithm to you.
rslt = minimize(self._criterion, x0, method='SLSQP',
constraints=constraints)
# Check for convergence
assert (rslt['success'] == True)
# Transformation of result.
x = rslt['x'] ** 2
# Type conversion
x = x.tolist()
# Quality Checks
integrity_checks('_choose_rational_out', x)
# Finishing
return x
def _choose(self, y, p1, p2):
""" Choose a random bundle on the budget line.
"""
# Antibugging
integrity_checks('_choose_random_in', y, p1, p2)
# Determine maximal consumption of good two
max_two = y / p2
# Initialize result container
x = [None, None]
# Select random bundle
x[1] = float(np.random.uniform(0, max_two))
x[0] = (y - x[1] * p2) / p1
# Quality Checks
integrity_checks('_choose_random_out', x)
# Finishing
return x
def _choose(self, y, p1, p2):
""" Choose a random bundle on the budget line.
"""
# Antibugging
integrity_checks('_choose_random_in', y, p1, p2)
# Determine maximal consumption of good two
max_two = y / p2
# Initialize result container
x = [None, None]
# Select random bundle
x[1] = float(np.random.uniform(0, max_two))
x[0] = (y - x[1] * p2) / p1
# Quality Checks
integrity_checks('_choose_random_out', x)
# Finishing
return x
