18182324

import math
 
increment = 0.1
startingPoint = [1, 1]
point1 = [1,5]
point2 = [6,4]
point3 = [5,2]
point4 = [2,1]
 
def distance(x1, y1, x2, y2):

18182324

i=0 
fitness = 0 
counter = 0 
while i < max_iter: 
   if counter > n_repeats: 
      STOP 
   Y = [ ]
   fit = [ ] 
   pairs = randompairsfrom (X) 
   for j in pairs: 

18182324

try: #If running in colab 
    import google.colab
    IN_COLAB = True
    %tensorflow_version 2.x
except:
    IN_COLAB = False
 
import tensorflow as tf
if (not tf.__version__.startswith('2')): #Checking if tf 2.0 is installed
    print('Please install tensorflow 2.0 to run this notebook')

18182324

try: #If running in colab 
    import google.colab
    IN_COLAB = True
    %tensorflow_version 2.x
except:
    IN_COLAB = False
 
import tensorflow as tf
if (not tf.__version__.startswith('2')): #Checking if tf 2.0 is installed
    print('Please install tensorflow 2.0 to run this notebook')

18182324

#Solution for model assumption. Different number of dollar signs. 
from scipy.stats import binom
#Define the considered numbers of dollar signs on the die (zero to six):
ndollar = np.asarray(np.linspace(0,6,7), dtype='int')
#Calculate corresponding probability of 2 $-signs in 10 throws
pdollar = binom.pmf(k=2, n=10, p=ndollar/6)
plt.stem(ndollar, pdollar)
plt.xlabel('Number of dollar signs on the dice')
plt.ylabel('Probability observing 2 dollar signs in 10 throws')
plt.title('Unnormalized likelihoods')

18182324

# Execute this cell to be sure to have a compatible TF (2.0) and TFP (0.8) version. 
# If you are bold you can skip this cell. 
try: #If running in colab 
  import google.colab
  !pip install tensorflow==2.0.0
  !pip install tensorflow_probability==0.8.0
except:
  print('Not running in colab')

try: #If running in colab 

18182324

#Reordering so x values are in increasiong order
order_idx_train=np.squeeze(x_train.argsort(axis=0))
x_train=x_train[order_idx_train]
y_train=y_train[order_idx_train]

order_idx_val=np.squeeze(x_val.argsort(axis=0))
x_val=x_val[order_idx_val]
y_val=y_val[order_idx_val]

order_idx_test=np.squeeze(x_test.argsort(axis=0))

18182324

try: #If running in colab 
    import google.colab
    IN_COLAB = True 
    %tensorflow_version 2.x
except:
    IN_COLAB = False
    
import tensorflow as tf
if (not tf.__version__.startswith('2')): #Checking if tf 2.0 is installed
    print('Please install tensorflow 2.0 to run this notebook')

18182324

theta=np.arange(0.05,1,0.05)
print(theta)
prior = 1/len(theta) #The normalizing constant of the prior

#Evaluate joint likelihood and unnormalized posterior at one specific #$\theta = 0.5$
dist = tfp.distributions.Bernoulli(probs=0.5) #one specific theta
print(np.prod(dist.prob(obs_data))) #joint likelihood
print(np.prod(dist.prob(obs_data))*prior) #unnormalized posterior

#Repeat the process for all thetas, range 0.05 - 0.95

18182324

py1_post = np.sum((res["theta"])*res["post"])
py0_post = 1.0 - py1_post
py0_post, py1_post

py1_prior = np.sum((res["theta"])*res["prior"]) 
py0_prior = 1 - py1_prior
py0_prior, py1_prior

#Plot posterior and prior
plt.figure(figsize=(16,12))