Lecture 74:- ML Titanic Challenge - 2 Data Analysis

Let's dive deeper into the data analysis for the Titanic challenge. We'll perform exploratory data analysis (EDA) to gain more insights into the dataset and understand how different features are related to the target variable "Survived."

pythonCopy code

import pandas as pd import numpy as np import matplotlib.pyplot as plt import seaborn as sns # Load the dataset train_df = pd.read_csv('train.csv') # Display basic information about the dataset print(train_df.info()) # Summary statistics of numerical features print(train_df.describe()) # Count of survivors print(train_df['Survived'].value_counts()) # Visualization: Bar plot of survival by gender sns.countplot(x='Sex', hue='Survived', data=train_df) plt.title('Survival by Gender') plt.show() # Visualization: Survival by passenger class sns.countplot(x='Pclass', hue='Survived', data=train_df) plt.title('Survival by Passenger Class') plt.show() # Visualization: Age distribution of passengers sns.histplot(data=train_df, x='Age', hue='Survived', multiple='stack') plt.title('Age Distribution of Passengers by Survival') plt.show() # Visualization: Fare distribution of passengers sns.histplot(data=train_df, x='Fare', hue='Survived', multiple='stack') plt.title('Fare Distribution of Passengers by Survival') plt.show() # Correlation heatmap corr_matrix = train_df.corr() sns.heatmap(corr_matrix, annot=True, cmap='coolwarm') plt.title('Correlation Heatmap') plt.show()

This code snippet demonstrates various data analysis techniques using Python and libraries like pandas, numpy, matplotlib, and seaborn. Here's what the code does:

1. Load the dataset using pd.read_csv().

2. Display basic information about the dataset using info().

3. Show summary statistics of numerical features using describe().

4. Count the number of survivors (0 = Not Survived, 1 = Survived).

5. Create bar plots to visualize survival by gender and passenger class.

6. Create histograms to visualize the age and fare distributions of passengers, stacked by survival status.

7. Generate a correlation heatmap to visualize the correlation between numerical features.

By analyzing the data and visualizing patterns, you can make informed decisions about feature selection, engineering, and model building. For example, from the visualizations, you might observe that survival rates vary by gender and passenger class, which suggests that these features could be important predictors for your machine learning model.

Remember that data analysis is an iterative process, and you might explore other features, create additional visualizations, and perform more in-depth analysis based on the insights you gain from this initial exploration.