First commit

Netlogo analysis.py

@@ -0,0 +1,126 @@
+# RP4 - Agent Based Model Analysis
+# Ben Mazin
+# 03/26/2024
+# School of Interdisciplinary Science, McMaster University
+
+# This script reads in the csv files from the Netlogo model and analyzes the data to determine if the chemo treatment is effective
+# It then saves the results to a csv file
+
+import pandas as pd
+from scipy import stats
+import matplotlib.pyplot as plt
+
+# Set the number of trials and initial hamsters
+number_of_trials = 200000
+initial_hamsters = 100
+
+# save path for the results
+save_path = '/Users/benmazin/Movies/'
+
+# Create an empty DataFrame with 4 columns
+all_data = []
+
+# Function to read CSV with inconsistent rows
+def read_inconsistent_csv(file_path, delimiter=','):
+    with open(file_path, 'r') as file:
+        lines = file.readlines()
+
+    # Determine the maximum number of columns by checking all rows
+    max_columns = max(len(line.split(delimiter)) for line in lines)
+
+    # Split each line into columns and pad with None if they are too short
+    data = [line.strip().split(delimiter) + [None] * (max_columns - len(line.split(delimiter))) for line in lines]
+
+    # Convert the list of lists into a DataFrame
+    df = pd.DataFrame(data, columns=range(max_columns))  # Set index to match column numbers
+    return df
+
+# Run for number_of_trials
+for i in range(1, number_of_trials + 1):
+
+    # Read in the csv file
+    file_path = f'{save_path}netlogo output/world{i}.csv'
+    data = read_inconsistent_csv(file_path, delimiter=',')
+
+    # Get the value of ChemoWork
+    chemowork = data.iat[6, 5]
+    if chemowork != '"true"' and chemowork != '"false"':
+        chemowork = data.iat[6, 3]
+
+    # Clean up the data
+    data = data.iloc[18:]
+    data = data.reset_index(drop=True)
+
+    # Get the total dead and max percent infected
+    total_dead = initial_hamsters - int(data.iloc[-1, 9].replace('"', ''))
+    max_percent_infect = int(data.iloc[2:, 1].str.replace('"', '').astype(int).max())
+
+    # Append the data to the proper list
+    if chemowork == '"true"':
+        all_data.append({
+            'Trial': i,
+            'ChemoTotalDead': total_dead,
+            'ChemoPercentInfect': max_percent_infect,
+            'NoChemoTotalDead': None,
+            'NoChemoPercentInfect': None
+        })
+
+    elif chemowork == '"false"':
+        all_data.append({
+            'Trial': i,
+            'ChemoTotalDead': None,
+            'ChemoPercentInfect': None,
+            'NoChemoTotalDead': total_dead,
+            'NoChemoPercentInfect': max_percent_infect
+        })
+
+    else:
+        print("Error: ChemoWork is not true or false")
+        print(i)
+
+#save dataframe to csv
+results = pd.DataFrame(all_data)
+
+# get the mean for each data column
+chemo_total_dead_mean = results['ChemoTotalDead'].mean()
+chemo_percent_infect_mean = results['ChemoPercentInfect'].mean()
+no_chemo_total_dead_mean = results['NoChemoTotalDead'].mean()
+no_chemo_percent_infect_mean = results['NoChemoPercentInfect'].mean()
+
+# run a t-test to determine if the means are significantly different for total dead and percent infect
+totalDeadTtest = stats.ttest_ind(results['ChemoTotalDead'], results['NoChemoTotalDead'] , nan_policy='omit')
+percentInfectTtest = stats.ttest_ind(results['ChemoPercentInfect'], results['NoChemoPercentInfect'], nan_policy='omit')
+
+# create a new dataframe with the means and t-test results
+resultSum = pd.DataFrame({
+    'Category': ['ChemoTotalDeadMean', 'ChemoPercentInfectMean', 'NoChemoTotalDeadMean', 'NoChemoPercentInfectMean', 'TotalDeadTtest', 'PercentInfectTtest'],
+    'Value': [chemo_total_dead_mean, chemo_percent_infect_mean, no_chemo_total_dead_mean, no_chemo_percent_infect_mean, totalDeadTtest, percentInfectTtest]
+})
+
+# Plot an individual distribution for each category and save to a png in the same directory
+save_path = save_path + 'netlogo results/'
+
+results['ChemoTotalDead'].plot.kde()
+plt.xlabel('Chemo Total Dead')
+plt.savefig(save_path + 'ChemoTotalDeadDist.png')
+plt.close()
+
+results['ChemoPercentInfect'].plot.kde()
+plt.xlabel('Chemo Percent Infect')
+plt.savefig(save_path + 'ChemoPercentInfectDist.png')
+plt.close()
+
+results['NoChemoTotalDead'].plot.kde()
+plt.xlabel('No Chemo Total Dead')
+plt.savefig(save_path + 'NoChemoTotalDeadDist.png')
+plt.close()
+
+results['NoChemoPercentInfect'].plot.kde()
+plt.xlabel('No Chemo Percent Infect')
+plt.savefig(save_path + 'NoChemoPercentInfectDist.png')
+plt.close()
+
+
+# save the summary and results to csv
+resultSum.to_csv(save_path + 'summary.csv', index=False)
+results.to_csv(save_path + 'results.csv', index=False)