Python for Data Analysis

Install Python 3.10+ and run: pip install pandas numpy matplotlib seaborn jupyter. Use Jupyter Notebooks for interactive data exploration — each cell runs independently, ideal for iterative analysis. Alternatively, Google Colab gives you a free hosted Jupyter environment with no setup.

pandas DataFrames are the core data structure. Key operations:

import pandas as pd

# Load data
df = pd.read_csv("data.csv")

# Inspect
df.head()          # First 5 rows
df.shape           # (rows, columns)
df.dtypes          # Column types
df.describe()      # Summary statistics
df.isnull().sum()  # Missing value counts

Real data is messy. Common cleaning operations:

# Drop duplicates
df.drop_duplicates(inplace=True)

# Fill missing values
df['salary'].fillna(df['salary'].median(), inplace=True)

# Drop columns with too many nulls
df.dropna(axis=1, thresh=len(df)*0.7, inplace=True)

# Convert types
df['date'] = pd.to_datetime(df['date'])
df['salary'] = df['salary'].astype(float)

The most powerful analysis patterns:

# Group by and aggregate
df.groupby('department')['salary'].agg(['mean', 'median', 'count'])

# Pivot table
pd.pivot_table(df,
  values='revenue',
  index='region',
  columns='product',
  aggfunc='sum',
  fill_value=0)

import matplotlib.pyplot as plt
import seaborn as sns

# Line chart
df.groupby('date')['revenue'].sum().plot()
plt.title("Revenue Over Time")
plt.show()

# Distribution
sns.histplot(df['salary'], bins=30)

# Correlation heatmap
sns.heatmap(df.select_dtypes('number').corr(),
  annot=True, cmap='coolwarm')