def P_data(data,m1,b1,m2,b2,σ):
x,y=data
μ=m1*x+b1
idx=where(x>=23)[0]
μ[idx]=m2*x[idx]+b2
distribution=Normal(μ,σ)
return sum(distribution(y))
def P_data(data, a0, a1, a2, σ):
x, y = data
μ = a0 + a1 * x + a2 * x * x # y ~ μ + random
distribution = Normal(μ, σ)
return sum(distribution(y))
def P_data(data, m, b, σ):
x, y = data
μ = m * x + b # y ~ μ + random
distribution = Normal(μ, σ)
return sum(distribution(y))
def P_data(data, m, b, σ):
x, y = data
μ = m * x + b # y ~ μ + random
distribution = Normal(μ, σ)
return sum(distribution(y))
# In[7]:
data = t, T
model = MCMCModel(
data,
P_data,
m=Normal(0, 10),
b=Uniform(-100, 100),
σ=Jeffreys(),
)
# In[8]:
for i in range(2):
model.run_mcmc(500)
model.plot_chains()
# In[9]:
model.BIC
# In[10]:
def P_data(data,C,p,Z,θ0,σ):
t,θ=data
μ=C*exp(-p*t)/(Z-p)+exp(-Z*t)*(θ0-C/(Z-p))
distribution=Normal(μ,σ)
return sum(distribution(θ))
# In[ ]:
# In[8]:
model=MCMCModel(sim,
Z=Normal(5,10),
)
model.run_mcmc(500,repeat=4)
model.plot_chains()
# In[9]:
model.plot_distributions()
# In[10]:
sim.run(0,11)
def P_data(data,μ,σ):
x=data
distribution=Normal(μ,σ)
return sum(distribution(data))
# In[26]:
def P_data(data,m,b,σ):
x,y=data
μ=m*x+b
distribution=Normal(μ,σ)
return sum(distribution(y))
# In[31]:
data=x-1960,y
model=MCMCModel(data,P_data,
m=Normal(0,10),
b=Uniform(-10,10),
σ=Jeffreys(),
)
# In[34]:
model.run_mcmc(500)
model.plot_chains()
# In[35]:
def P_data(data, m, b, σ):
x, y = data
μ = m * x + b
distribution = Normal(μ, σ)
return sum(distribution(y))
# In[8]:
data = x, y
model = MCMCModel(
data,
P_data,
m=Normal(0, 10),
b=Uniform(-10, 10),
σ=Jeffreys(),
)
# In[9]:
for i in range(2):
model.run_mcmc(500)
model.plot_chains()
# In[17]:
model.plot_distributions()
# In[ ]:
def P_data(data, μ):
x = data
σ = 0.1
distribution = Normal(μ, σ)
return sum(distribution(x))
plot(y, m, '-o')
# In[32]:
def P_data(data, μ):
x = data
σ = 0.1
distribution = Normal(μ, σ)
return sum(distribution(x))
# In[39]:
model = MCMCModel(m, P_data, μ=Normal(0, 20))
# In[40]:
model.run_mcmc(500)
model.plot_chains()
# In[41]:
model.plot_distributions()
# In[42]:
def P_data(data, μ, σ):
x = data