Created in May 2024
PDF: paulgalea.com/Projects/Titanic_Survival/Visualisation.pdf

/* ..................... */
/* ........TOOLS........ */
/* ..................... */

Python
Adobe Illustrator

/* ..................... */
/* ....PYTHON SCRIPT.... */
/* ..................... */

import pandas as pd  
from sklearn.tree import DecisionTreeClassifier, plot_tree   
from sklearn.ensemble import RandomForestClassifier   
from sklearn.linear_model import LogisticRegression  
from sklearn.model_selection import cross_val_score 
import matplotlib.pyplot as plt  
   
 
# Data  
d = pd.read_excel(r'DATA LOCATION')  
d = d.dropna(subset=['Age']).assign(Survived=d['Survived'].map({'N':0,'Y':1}), Sex=d['Sex'].map({'M':1,'F':0}))   
d['Age'] = pd.cut(d['Age'], bins=[0,10,40,float('inf')], labels=['Age 0-9','Age 10-39','Age 40+'], right=False) 
print(d.groupby(['Sex','Position','Age'], observed=False)['Survived'].value_counts().unstack())  
d = pd.concat([d,pd.get_dummies(d[['Position','Age']], prefix='', prefix_sep='').astype(int)], axis=1)   
  
  
# Model
features, target = d.drop(columns=['Name','Age','Position','Survived']), d['Survived']
models, results, cv, random_state, class_weight = {}, {}, 5, 0, 'balanced'

models = {
    'Decision Tree':DecisionTreeClassifier(class_weight=class_weight, max_depth=3, random_state=random_state),
    'Random Forest':RandomForestClassifier(class_weight=class_weight, random_state=random_state),
    'Logistic Regression':LogisticRegression(class_weight=class_weight, random_state=random_state)}

for name, model in models.items():
    accuracy = cross_val_score(model, features, target, cv=cv, scoring='accuracy').mean()
    precision = cross_val_score(model, features, target, cv=cv, scoring='precision').mean()
    recall = cross_val_score(model, features, target, cv=cv, scoring='recall').mean()
    f1 = cross_val_score(model, features, target, cv=cv, scoring='f1').mean()
    results[name] = {'Name':name,'Accuracy':accuracy,'Precision':precision,'Recall':recall,'F1':f1}

print('\nFeatures:', ', '.join(features.columns), '\n\n', round(pd.DataFrame(results.values()),2))


# Visual
plot_tree(models['Decision Tree'].fit(features, target), feature_names=features.columns,
          class_names=['Lost','Saved'], filled=True, fontsize=7)
plt.show()

/* ..................... */
/* .......SOURCES....... */
/* ..................... */

Viewed online May 2024:
1. historyonthenet.com/the-titanic-why-did-people-believe-titanic-was-unsinkable
2. britannica.com/topic/titanic
3. deseret.com/23602652/titanic-lifeboats
4. encyclopedia-titanica.org/explorer (used for passenger count data)
5. titanic.fandom.com/wiki/women_and_children_first
6. espace.library.uq.edu.au/view/uq:152940
7. encyclopedia-titanica.org/titanic-deckplans
8. encyclopedia-titanica.org/titanics-aft-lifeboats.html
9. historyonthenet.com/last-night-titanic-cooks