petermenegay: Create a 2x2 array of numbers with the following values, [0, .10, .05], [.10, 0, .10], and [0, .05, 0]

GitHub Copilot: ### [policy_fixed_point.py](file:///C:/A/PeerVerity/policy_system_ai/policy_fixed_point.py)

Add a 2x2 array of numbers with the specified values.

```python


# ...existing code...

import numpy as np

array_2x2 = np.array([[0, 0.10, 0.05], [0.10, 0, 0.10], [0, 0.05, 0]])

# ...existing code...
```

Made changes.

petermenegay: Add print statements for all the EFF values

GitHub Copilot: ### [policy_fixed_point.py](file:///C:/A/PeerVerity/policy_system_ai/policy_fixed_point.py)

Add print statements for all the EFF values.

```python


# ...existing code...

print(EFF[0][0])
print(EFF[0][1])
print(EFF[0][2])
print(EFF[1][0])
print(EFF[1][1])
print(EFF[1][2])
print(EFF[2][0])
print(EFF[2][1])
print(EFF[2][2])

# ...existing code...
```

Made changes.

petermenegay: Do the calculation for IMP over and over until IMP converges to some value.

GitHub Copilot: ### [policy_fixed_point.py](file:///C:/A/PeerVerity/policy_system_ai/policy_fixed_point.py)

Modify the code to perform the calculation for IMP until it converges.

```python


# ...existing code...

tolerance = 1e-6
max_iterations = 1000

for _ in range(max_iterations):
    new_IMP = IWM + np.dot(EFFtrans, IMP)
    if np.allclose(new_IMP, IMP, atol=tolerance):
        break
    IMP = new_IMP

print('Converged IMP = ', IMP)

# ...existing code...
```

Made changes.

petermenegay: Prevent IMP from going over 100. That is, if it is greater than 100, make it equal to 100.

GitHub Copilot: ### [policy_fixed_point.py](file:///C:/A/PeerVerity/policy_system_ai/policy_fixed_point.py)

Modify the code to prevent IMP from going over 100.

```python


# ...existing code...

tolerance = 1e-6
max_iterations = 1000
IMPmax_value = 100.0

for _ in range(max_iterations):
    new_IMP = IWM + np.dot(EFF, IMP)
    new_IMP = np.clip(new_IMP, None, IMPmax_value)  # Ensure IMP does not exceed 100
    if np.allclose(new_IMP, IMP, atol=tolerance):
        break
    IMP = new_IMP

print('Converged IMP = ', IMP)

# ...existing code...
```

Made changes.

petermenegay: Below where it says #Design of Experiments, set up the code to vary the EXPEND values to see how VOTHAPtot changes. Vary EXPEND between 0 and CTA.

GitHub Copilot: ### [policy_fixed_point.py](file:///C:/A/PeerVerity/policy_system_ai/policy_fixed_point.py)

Set up the code to vary the EXPEND values between 0 and CTA to see how VOTHAPtot changes.

```python


# ...existing code...

#Design of Experiments
# Vary the EXPEND values and see how VOTHAPtot changes
EXPEND_ranges = [np.linspace(0, cta, num=5) for cta in CTA]

for exp1 in EXPEND_ranges[0]:
    for exp2 in EXPEND_ranges[1]:
        for exp3 in EXPEND_ranges[2]:
            EXPEND_test = np.array([exp1, exp2, exp3])
            VOTHAPtot = voter_happiness(P, Pb, EFF, IMPmax, CTA, EXPEND_test, EXPENDnom)
            print(f'EXPEND = {EXPEND_test}, VOTHAPtot = {VOTHAPtot}')

# ...existing code...
```

Made changes.

petermenegay: Comment out all the print statements except the last one

GitHub Copilot: ### [policy_fixed_point.py](file:///C:/A/PeerVerity/policy_system_ai/policy_fixed_point.py)

Comment out all the print statements except the last one.

```python


# ...existing code...

def voter_happiness(P, Pb, EFF, IMPmax, CTA, EXPEND, EXPENDnom):
    #Calculate the total priority and the cost per point for each policy
    Ptot = np.sum(P) + Pb #Total priority (sum of policy priorities and budget priority)
    CPP = CTA/IMPmax #Cost per point for each policy
    # print('CPP = ', CPP)
    EXPENDtot = EXPENDb = np.sum(EXPEND) #Total expenditure on policies
    #EXPENDtot = EXPENDb = 0.0
    IWM = EXPEND/CPP #Improvement with money -- how much improvement is achieved solely through expenditure
    # print('IWM = ', IWM)
    #The voter is most happy if no money is spent on any policy, thus giving IMPmax for the budget
    # beyond that happiness decreases linearly with expenditure
    IMPb = IMPmax - (IMPmax/EXPENDnom)*EXPENDtot #Improvement from budget
    IMP = IWM #initial IMP
    tolerance = 1e-6
    max_iterations = 1000
    for _ in range(max_iterations):
        new_IMP = IWM + np.dot(EFF, IMP)
        new_IMP = np.clip(new_IMP, None, IMPmax)  # Ensure IMP does not exceed 100
        if np.allclose(new_IMP, IMP, atol=tolerance):
            break
        IMP = new_IMP
    # print('Converged IMP = ', IMP)
    VOTHAP = IMP*P/Ptot #Voter's happiness from each policy
    # print('VOTHAP = ', VOTHAP)
    # print('IMPb = ', IMPb)
    VOTHAPb = IMPb*Pb/Ptot #Voter's happiness from budget
    # print('VOTHAPb = ', VOTHAPb)
    VOTHAPtot = np.sum(VOTHAP) + VOTHAPb
    # print('VOTHAPtot = ', VOTHAPtot)
    return VOTHAPtot

# ...existing code...

print('VOTHAPtot = ', VOTHAPtot)

#Design of Experiments
# Vary the EXPEND values and see how VOTHAPtot changes
EXPEND_ranges = [np.linspace(0, cta, num=5) for cta in CTA]

for exp1 in EXPEND_ranges[0]:
    for exp2 in EXPEND_ranges[1]:
        for exp3 in EXPEND_ranges[2]:
            EXPEND_test = np.array([exp1, exp2, exp3])
            VOTHAPtot = voter_happiness(P, Pb, EFF, IMPmax, CTA, EXPEND_test, EXPENDnom)
            print(f'EXPEND = {EXPEND_test}, VOTHAPtot = {VOTHAPtot}')

# ...existing code...
```

Made changes.

petermenegay: For the last print statement, print out just the values with tabs to separate them. Print the headers of the table at the top, eg EXPEND[0], EXPEND[1], etc and VOTHAPtot

GitHub Copilot: ### [policy_fixed_point.py](file:///C:/A/PeerVerity/policy_system_ai/policy_fixed_point.py)

Modify the last print statement to print out just the values with tabs to separate them, and print the headers of the table at the top.

```python


# ...existing code...

#Design of Experiments
# Vary the EXPEND values and see how VOTHAPtot changes
EXPEND_ranges = [np.linspace(0, cta, num=5) for cta in CTA]

# Print headers
print("EXPEND[0]\tEXPEND[1]\tEXPEND[2]\tVOTHAPtot")

for exp1 in EXPEND_ranges[0]:
    for exp2 in EXPEND_ranges[1]:
        for exp3 in EXPEND_ranges[2]:
            EXPEND_test = np.array([exp1, exp2, exp3])
            VOTHAPtot = voter_happiness(P, Pb, EFF, IMPmax, CTA, EXPEND_test, EXPENDnom)
            print(f'{exp1}\t{exp2}\t{exp3}\t{VOTHAPtot}')

# ...existing code...
```

Made changes.

petermenegay: Also add the total expenditure to the printed table

GitHub Copilot: ### [policy_fixed_point.py](file:///C:/A/PeerVerity/policy_system_ai/policy_fixed_point.py)

Add the total expenditure to the printed table.

```python


# ...existing code...

#Design of Experiments
# Vary the EXPEND values and see how VOTHAPtot changes
EXPEND_ranges = [np.linspace(0, cta, num=5) for cta in CTA]

# Print headers
print("EXPEND[0]\tEXPEND[1]\tEXPEND[2]\tTotal EXPEND\tVOTHAPtot")

for exp1 in EXPEND_ranges[0]:
    for exp2 in EXPEND_ranges[1]:
        for exp3 in EXPEND_ranges[2]]:
            EXPEND_test = np.array([exp1, exp2, exp3])
            total_expend = np.sum(EXPEND_test)
            VOTHAPtot = voter_happiness(P, Pb, EFF, IMPmax, CTA, EXPEND_test, EXPENDnom)
            print(f'{exp1}\t{exp2}\t{exp3}\t{total_expend}\t{VOTHAPtot}')

# ...existing code...
```

Made changes.

petermenegay: Write out the table to a file

GitHub Copilot: ### [policy_fixed_point.py](file:///C:/A/PeerVerity/policy_system_ai/policy_fixed_point.py)

Write out the table to a file.

```python


# ...existing code...

#Design of Experiments
# Vary the EXPEND values and see how VOTHAPtot changes
EXPEND_ranges = [np.linspace(0, cta, num=5) for cta in CTA]

# Open file to write the table
with open('vothap_results.txt', 'w') as file:
    # Write headers
    file.write("EXPEND[0]\tEXPEND[1]\tEXPEND[2]\tTotalEXPEND\tVOTHAPtot\n")
    
    for exp1 in EXPEND_ranges[0]:
        for exp2 in EXPEND_ranges[1]:
            for exp3 in EXPEND_ranges[2]:
                EXPEND_test = np.array([exp1, exp2, exp3])
                total_expend = np.sum(EXPEND_test)
                VOTHAPtot = voter_happiness(P, Pb, EFF, IMPmax, CTA, EXPEND_test, EXPENDnom)
                file.write(f'{exp1}\t{exp2}\t{exp3}\t{total_expend}\t{VOTHAPtot}\n')

# ...existing code...
```

Made changes.

petermenegay: In addition to printing, write out the results to a file

GitHub Copilot: ### [policy_fixed_point.py](file:///C:/A/PeerVerity/policy_system_ai/policy_fixed_point.py)

In addition to printing, write out the results to a file.

```python


# ...existing code...

#Design of Experiments
# Vary the EXPEND values and see how VOTHAPtot changes
EXPEND_ranges = [np.linspace(0, cta, num=5) for cta in CTA]

# Open file to write the table
with open('vothap_results.txt', 'w') as file:
    # Write headers
    file.write("EXPEND[0]\tEXPEND[1]\tEXPEND[2]\tTotalEXPEND\tVOTHAPtot\n")
    print("EXPEND[0]\tEXPEND[1]\tEXPEND[2]\tTotalEXPEND\tVOTHAPtot")
    
    for exp1 in EXPEND_ranges[0]:
        for exp2 in EXPEND_ranges[1]:
            for exp3 in EXPEND_ranges[2]:
                EXPEND_test = np.array([exp1, exp2, exp3])
                total_expend = np.sum(EXPEND_test)
                VOTHAPtot = voter_happiness(P, Pb, EFF, IMPmax, CTA, EXPEND_test, EXPENDnom)
                result_line = f'{exp1}\t{exp2}\t{exp3}\t{total_expend}\t{VOTHAPtot}\n'
                file.write(result_line)
                print(result_line.strip())

# ...existing code...
```

Made changes.

petermenegay: Instead of varying EXPEND in a full factorial manner, vary it randomly. Use 200 points. 

GitHub Copilot: ### [policy_fixed_point.py](file:///C:/A/PeerVerity/policy_system_ai/policy_fixed_point.py)

Vary EXPEND randomly using 200 points.

```python


# ...existing code...

#Design of Experiments
# Vary the EXPEND values randomly and see how VOTHAPtot changes
num_points = 200
EXPEND_ranges = [np.random.uniform(0, cta, num_points) for cta in CTA]

# Open file to write the table
with open('vothap_results.txt', 'w') as file:
    # Write headers
    file.write("EXPEND[0]\tEXPEND[1]\tEXPEND[2]\tTotalEXPEND\tVOTHAPtot\n")
    print("EXPEND[0]\tEXPEND[1]\tEXPEND[2]\tTotalEXPEND\tVOTHAPtot")
    
    for i in range(num_points):
        exp1 = EXPEND_ranges[0][i]
        exp2 = EXPEND_ranges[1][i]
        exp3 = EXPEND_ranges[2][i]
        EXPEND_test = np.array([exp1, exp2, exp3])
        total_expend = np.sum(EXPEND_test)
        VOTHAPtot = voter_happiness(P, Pb, EFF, IMPmax, CTA, EXPEND_test, EXPENDnom)
        result_line = f'{exp1}\t{exp2}\t{exp3}\t{total_expend}\t{VOTHAPtot}\n'
        file.write(result_line)
        print(result_line.strip())

# ...existing code...
```

Made changes.

petermenegay: Add a plot of total expenditure on the x-axis and total voter happiness on the y-axis.

GitHub Copilot: ### [policy_fixed_point.py](file:///C:/A/PeerVerity/policy_system_ai/policy_fixed_point.py)

Add a plot of total expenditure on the x-axis and total voter happiness on the y-axis.

```python


# ...existing code...

import matplotlib.pyplot as plt

#Design of Experiments
# Vary the EXPEND values randomly and see how VOTHAPtot changes
num_points = 200
EXPEND_ranges = [np.random.uniform(0, cta, num_points) for cta in CTA]

total_expenditures = []
vothap_totals = []

# Open file to write the table
with open('vothap_results.txt', 'w') as file:
    # Write headers
    file.write("EXPEND[0]\tEXPEND[1]\tEXPEND[2]\tTotalEXPEND\tVOTHAPtot\n")
    print("EXPEND[0]\tEXPEND[1]\tEXPEND[2]\tTotalEXPEND\tVOTHAPtot")
    
    for i in range(num_points):
        exp1 = EXPEND_ranges[0][i]
        exp2 = EXPEND_ranges[1][i]
        exp3 = EXPEND_ranges[2][i]
        EXPEND_test = np.array([exp1, exp2, exp3])
        total_expend = np.sum(EXPEND_test)
        VOTHAPtot = voter_happiness(P, Pb, EFF, IMPmax, CTA, EXPEND_test, EXPENDnom)
        result_line = f'{exp1}\t{exp2}\t{exp3}\t{total_expend}\t{VOTHAPtot}\n'
        file.write(result_line)
        print(result_line.strip())
        
        total_expenditures.append(total_expend)
        vothap_totals.append(VOTHAPtot)

# Plotting the results
plt.scatter(total_expenditures, vothap_totals)
plt.xlabel('Total Expenditure')
plt.ylabel('Total Voter Happiness')
plt.title('Total Expenditure vs Total Voter Happiness')
plt.grid(True)
plt.show()

# ...existing code...
```

Made changes.

petermenegay: Add code to find the Pareto frontier of the results, that is the points that represent the highest voter happiness for each cost. Write the pareto frontier points out to a file and plot them on the same graph as the total results.

GitHub Copilot: ### [policy_fixed_point.py](file:///C:/A/PeerVerity/policy_system_ai/policy_fixed_point.py)

Add code to find the Pareto frontier of the results, write the Pareto frontier points to a file, and plot them on the same graph as the total results.

```python


# ...existing code...

def find_pareto_frontier(costs, values):
    sorted_indices = np.argsort(costs)
    pareto_costs = []
    pareto_values = []
    max_value = -np.inf
    for idx in sorted_indices:
        cost = costs[idx]
        value = values[idx]
        if value > max_value:
            pareto_costs.append(cost)
            pareto_values.append(value)
            max_value = value
    return pareto_costs, pareto_values

#Design of Experiments
# Vary the EXPEND values randomly and see how VOTHAPtot changes
num_points = 200
EXPEND_ranges = [np.random.uniform(0, cta, num_points) for cta in CTA]

total_expenditures = []
vothap_totals = []

# Open file to write the table
with open('vothap_results.txt', 'w') as file:
    # Write headers
    file.write("EXPEND[0]\tEXPEND[1]\tEXPEND[2]\tTotalEXPEND\tVOTHAPtot\n")
    print("EXPEND[0]\tEXPEND[1]\tEXPEND[2]\tTotalEXPEND\tVOTHAPtot")
    
    for i in range(num_points):
        exp1 = EXPEND_ranges[0][i]
        exp2 = EXPEND_ranges[1][i]
        exp3 = EXPEND_ranges[2][i]
        EXPEND_test = np.array([exp1, exp2, exp3])
        total_expend = np.sum(EXPEND_test)
        VOTHAPtot = voter_happiness(P, Pb, EFF, IMPmax, CTA, EXPEND_test, EXPENDnom)
        result_line = f'{exp1}\t{exp2}\t{exp3}\t{total_expend}\t{VOTHAPtot}\n'
        file.write(result_line)
        print(result_line.strip())
        
        total_expenditures.append(total_expend)
        vothap_totals.append(VOTHAPtot)

# Find Pareto frontier
pareto_costs, pareto_values = find_pareto_frontier(total_expenditures, vothap_totals)

# Write Pareto frontier to a file
with open('pareto_frontier.txt', 'w') as pareto_file:
    pareto_file.write("TotalEXPEND\tVOTHAPtot\n")
    for cost, value in zip(pareto_costs, pareto_values):
        pareto_file.write(f'{cost}\t{value}\n')

# Plotting the results
plt.scatter(total_expenditures, vothap_totals, label='Total Results')
plt.plot(pareto_costs, pareto_values, color='red', label='Pareto Frontier')
plt.xlabel('Total Expenditure')
plt.ylabel('Total Voter Happiness')
plt.title('Total Expenditure vs Total Voter Happiness')
plt.legend()
plt.grid(True)
plt.show()

# ...existing code...
```

Made changes.

petermenegay: Instead of a red line to connect the Pareto frontier points, just identify the Pareto points using a different color.

GitHub Copilot: ### [policy_fixed_point.py](file:///C:/A/PeerVerity/policy_system_ai/policy_fixed_point.py)

Identify the Pareto frontier points using a different color instead of connecting them with a red line.

```python


# ...existing code...

# Plotting the results
plt.scatter(total_expenditures, vothap_totals, label='Total Results')
plt.scatter(pareto_costs, pareto_values, color='red', label='Pareto Frontier')
plt.xlabel('Total Expenditure')
plt.ylabel('Total Voter Happiness')
plt.title('Total Expenditure vs Total Voter Happiness')
plt.legend()
plt.grid(True)
plt.show()

# ...existing code...
```

Made changes.