Skip to content

Commit

Permalink
fixed path to comparisons
Browse files Browse the repository at this point in the history
  • Loading branch information
@Almog-David
Almog-David committed Aug 2, 2024
1 parent ecb10c3 commit 33f73e0
Show file tree
Hide file tree
Showing 2 changed files with 144 additions and 20 deletions.
50 changes: 30 additions & 20 deletions experiments/firefighter_problem/comparisons.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -118,7 +118,7 @@ def Compare_NonSpread():
This function runs multiple experiments on randomly generated layered networks
and plots the results comparing the budget used by different algorithms.
"""
ex1 = experiments_csv.Experiment("./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/", "non_spreading.csv", backup_folder=None)
ex1 = experiments_csv.Experiment("./experiments/", "non_spreading.csv", backup_folder=None)
ex1.clear_previous_results() # to clear previous experiments

input_ranges = {
Expand All @@ -144,7 +144,7 @@ def multiple_runs(runs=30):
ex1.run_with_time_limit(multiple_runs, input_ranges={}, time_limit=0.9)

# Preprocess the DataFrame to extract numeric budget values
results_csv_file = "./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/non_spreading_minbudget.csv"
results_csv_file = "./experiments/non_spreading_minbudget.csv"
results = pd.read_csv(results_csv_file)

# Extract the numeric budget from the 'Budget' column
Expand All @@ -164,7 +164,7 @@ def extract_budget_numeric(budget):
results = results.dropna(subset=['Budget_numeric'])

# Save the preprocessed DataFrame to a temporary CSV file
preprocessed_csv_file = "./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/non_spreading_minbudget_preprocessed.csv"
preprocessed_csv_file = "./experiments/non_spreading_minbudget_preprocessed.csv"
results.to_csv(preprocessed_csv_file, index=False)

print("\n DataFrame-NonSpread: \n", results)
Expand All @@ -177,7 +177,7 @@ def extract_budget_numeric(budget):
y_field="Budget_numeric",
z_field="algorithm",
mean=True,
save_to_file="./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/non_spreading.png"
save_to_file="./experiments/non_spreading.png"
)

print("\n DataFrame-NonSpread: \n", ex1.dataFrame)
Expand All @@ -189,7 +189,7 @@ def Compare_SpreadingMaxSave():
This function runs multiple experiments on randomly generated directed graphs
and plots the results comparing the number of nodes saved by different algorithms.
"""
ex2 = experiments_csv.Experiment("./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/", "spreading_maxsave.csv", backup_folder=None)
ex2 = experiments_csv.Experiment("./experiments/", "spreading_maxsave.csv", backup_folder=None)
ex2.clear_previous_results() # to clear previous experiments

input_ranges = {
Expand Down Expand Up @@ -224,7 +224,7 @@ def multiple_runs(runs=10):

## DATA ISSUE WE HAD SO THIS IS A FIX ##
# Load the results
results_csv_file = "./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/spreading_maxsave.csv"
results_csv_file = "./experiments/spreading_maxsave.csv"
results = pd.read_csv(results_csv_file)

# Ensure 'algorithm' column is of type string
Expand All @@ -245,7 +245,7 @@ def multiple_runs(runs=10):
results['Nodes_Saved'] = results['Nodes_Saved'].astype(int)

# Save the cleaned DataFrame to a new CSV file (optional, for debugging)
cleaned_csv_file = "./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/spreading_maxsave_preprocessed.csv"
cleaned_csv_file = "./experiments/spreading_maxsave_preprocessed.csv"
results.to_csv(cleaned_csv_file, index=False)

# Plot the results using the cleaned DataFrame
Expand All @@ -261,7 +261,7 @@ def multiple_runs(runs=10):
sharex=True,
sharey=True,
mean=True,
save_to_file="./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/spreading_maxsave_budget.png"
save_to_file="./experiments/spreading_maxsave_budget.png"
)

multi_plot_results(
Expand All @@ -276,7 +276,7 @@ def multiple_runs(runs=10):
sharex=True,
sharey=True,
mean=True,
save_to_file="./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/spreading_maxsave_100_edge_prob.png"
save_to_file="./experiments/spreading_maxsave_100_edge_prob.png"
)

multi_plot_results(
Expand All @@ -291,7 +291,7 @@ def multiple_runs(runs=10):
sharex=True,
sharey=True,
mean=True,
save_to_file="./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/spreading_maxsave_200_edge_prob.png"
save_to_file="./experiments/spreading_maxsave_200_edge_prob.png"
)

multi_plot_results(
Expand All @@ -306,7 +306,7 @@ def multiple_runs(runs=10):
sharex=True,
sharey=True,
mean=True,
save_to_file="./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/spreading_maxsave_400_edge_prob.png"
save_to_file="./experiments/spreading_maxsave_400_edge_prob.png"
)

print("\n DataFrame-NonSpread: \n", ex2.dataFrame)
Expand All @@ -318,7 +318,7 @@ def Compare_SpreadingMinBudget():
This function runs multiple experiments on randomly generated directed graphs
and plots the results comparing the budget used by different algorithms.
"""
ex3 = experiments_csv.Experiment("./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/", "spreading_minbudget.csv", backup_folder=None)
ex3 = experiments_csv.Experiment("./experiments/", "spreading_minbudget.csv", backup_folder=None)
ex3.clear_previous_results() # to clear previous experiments

input_ranges = {
Expand All @@ -328,7 +328,7 @@ def Compare_SpreadingMinBudget():
node_counts = [100, 200, 400]
edge_probabilities = [0.1, 0.5, 0.8]

def multiple_runs(runs=30):
def multiple_runs(runs=10):
for num_nodes in node_counts:
for edge_prob in edge_probabilities:
graph = generate_random_DiGraph(num_nodes=num_nodes, edge_probability=edge_prob, seed=None)
Expand All @@ -340,16 +340,16 @@ def multiple_runs(runs=30):
targets = random.sample(nodes, num_targets)
for algorithm in input_ranges["algorithm"]:
start_time = perf_counter()
result = runner_spreading(algorithm, graph, source, targets)
result = runner_spreading(algorithm, graph, None, source, targets)
runtime = perf_counter() - start_time
ex3.add({**{"algorithm": algorithm.__name__, "runtime": runtime, "graph_nodes": len(graph.nodes)}, **result})
ex3.add({**{"algorithm": algorithm.__name__, "runtime": runtime, "graph_nodes": num_nodes, "edge_probability": edge_prob}, **result})
return {"status": "completed"}

# Set a time limit for the entire batch run
ex3.run_with_time_limit(multiple_runs, input_ranges={}, time_limit=0.9)

# Preprocess the DataFrame to extract numeric budget values
results_csv_file = "./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/spreading_minbudget.csv"
results_csv_file = "./experiments/spreading_minbudget.csv"
results = pd.read_csv(results_csv_file)

# Extract the numeric budget from the 'Budget' column
Expand All @@ -369,7 +369,7 @@ def extract_budget_numeric(budget):
results = results.dropna(subset=['Budget_numeric'])

# Save the preprocessed DataFrame to a temporary CSV file
preprocessed_csv_file = "./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/spreading_minbudget_preprocessed.csv"
preprocessed_csv_file = "./experiments/spreading_minbudget_preprocessed.csv"
results.to_csv(preprocessed_csv_file, index=False)

print("\n DataFrame-NonSpread: \n", results)
Expand All @@ -382,7 +382,17 @@ def extract_budget_numeric(budget):
y_field="Budget_numeric",
z_field="algorithm",
mean=True,
save_to_file="./networkz/algorithms/approximation/tests/test_firefighter_problem/comparisons/spreading_minbudget.png"
save_to_file="./experiments/spreading_minbudget.png"
)

single_plot_results(
results_csv_file=preprocessed_csv_file,
filter={"edge_probability":0.1},
x_field="graph_nodes",
y_field="Budget_numeric",
z_field="algorithm",
mean=True,
save_to_file="./experiments/spreading_minbudget_edge.png"
)

print("\n DataFrame-NonSpread: \n", ex3.dataFrame)
Expand All @@ -393,5 +403,5 @@ def extract_budget_numeric(budget):
setup_global_logger(level=logging.DEBUG)

#Compare_NonSpread()
# Compare_SpreadingMinBudget()
Compare_SpreadingMaxSave()
Compare_SpreadingMinBudget()
#Compare_SpreadingMaxSave()
Loading

0 comments on commit 33f73e0

Please sign in to comment.