sonarkode.blog

Hello, everyone! In this post, we’re going to explore the fundamentals of data analysis with Python.
We’ll use three of the most essential libraries in the data science ecosystem: Pandas, NumPy, and Matplotlib. By the end, you’ll be able to load, manipulate, and visualize datasets directly in a Jupyter Notebook inside Visual Studio Code.

Why Pandas, NumPy, and Matplotlib?

• NumPy: Provides fast array operations and numerical computing.
• Pandas: Built on top of NumPy, designed for working with tabular data (rows and columns).
• Matplotlib: A powerful plotting library to visualize data with charts and graphs.

Together, these three form the backbone of most Python data science projects.

Step 1: Setting Up the Environment

Make sure Python and VS Code are installed. Then, install the required libraries:

pip install numpy pandas matplotlib jupyter

Create a new file in VS Code and save it as analysis.ipynb. VS Code will render it as a Jupyter Notebook with runnable cells.

Step 2: Create a simple dataset with Pandas

Open a new code cell and import Pandas. Then create a small DataFrame:

import pandas as pd

data = {
    "Name": ["Alice", "Bob", "Charlie", "Diana", "Edward"],
    "Age": [25, 30, 35, 40, 29],
    "Salary": [50000, 60000, 70000, 80000, 55000]
}

df = pd.DataFrame(data)
df.head()

Step 3: Basic analysis with NumPy and Pandas

Convert columns to NumPy arrays for numeric ops, and use Pandas methods for quick stats:

import numpy as np

ages = df["Age"].to_numpy()
salaries = df["Salary"].to_numpy()

print("Average age:", np.mean(ages))
print("Max salary:", np.max(salaries))

# Pandas quick summary
df.describe()

df.describe() gives count, mean, std, min/max and quartiles — very handy for initial EDA.

Step 4: Data cleaning examples (short)

Real datasets need cleaning. Examples:

# drop rows with missing Salary
df_clean = df.dropna(subset=["Salary"])

# convert a column type
df_clean["Salary"] = df_clean["Salary"].astype(int)

# filter rows
df_filtered = df_clean[df_clean["Age"] >= 30]

Step 5: Visualization with Matplotlib

Plot a bar chart and a scatter/line to understand distributions and relationships.

import matplotlib.pyplot as plt

# Bar chart: Salary by Name
plt.figure(figsize=(8, 4))
plt.bar(df["Name"], df["Salary"])
plt.title("Salary Comparison")
plt.xlabel("Employee")
plt.ylabel("Salary")
plt.show()

# Scatter / line: Age vs Salary
plt.figure(figsize=(8, 4))
plt.plot(df["Age"], df["Salary"], marker="o")
plt.title("Age vs Salary")
plt.xlabel("Age")
plt.ylabel("Salary")
plt.grid(True)
plt.show()

Charts in the notebook show up inline — ideal for iterating quickly.

Step 6: Combining everything: small analysis pipeline

# 1. Load or create data (here we already have df)
# 2. Quick EDA
print(df.describe())

# 3. Feature engineering (example: salary per age)
df["Salary_per_Age"] = df["Salary"] / df["Age"]

# 4. Visualize the new metric
plt.figure(figsize=(8,4))
plt.bar(df["Name"], df["Salary_per_Age"])
plt.title("Salary per Age")
plt.xlabel("Employee")
plt.ylabel("Salary / Age")
plt.show()

This pattern is; load → clean → explore → visualize → iterate — is the core loop of data analysis.