Pandas is a foundational library for data manipulation and analysis in Python. This guide covers core constructs—Series and DataFrame—along with essential operations including indexing, sorting, and structural inspection.

Installation and Setup

Install the library via pip:

pip install pandas

Import conventionally to ensure clean, readable code:

import pandas as pd
import numpy as np

Core Data Structures

Create a one-dimensional labeled array using Series:

sample_series = pd.Series([1, 3, 6, np.nan, 44, 1])
print(sample_series)

Generate a sequence of dates for time-aware indexing:

timeline = pd.date_range("2023-01-01", periods=6, freq="D")
print(timeline)

Construct a two-dimensional DataFrame with custom row indices and column labels:

data_matrix = np.random.randn(6, 4)
df_main = pd.DataFrame(data_matrix, index=timeline, columns=["col_x", "col_y", "col_z", "col_w"])
print(df_main)

Initialize a DataFrame from a NumPy array without explicit labels:

basic_df = pd.DataFrame(np.arange(12).reshape(3, 4))
print(basic_df)

Build a DataFrame using heterogeneous data types via dictionary:

mixed_df = pd.DataFrame({
    "alpha": 1.0,
    "beta": pd.Timestamp("2022-05-10"),
    "gamma": pd.Series([2] * 4, dtype="int64"),
    "delta": np.array([7] * 4, dtype="int32"),
    "epsilon": pd.Categorical(["low", "high", "low", "high"]),
    "zeta": "static"
})
print(mixed_df)

Structural Inspection & Transformation

Inspect metadata and content:

• mixed_df.dtypes — reveals per-column data types
• mixed_df.index — returns row labels
• mixed_df.columns — returns column names
• mixed_df.values — returns underlying NumPy array
• mixed_df.describe() — computes summary statistics (numeric columns only)
• mixed_df.T — transposes rows and columns

Sorting Operations

Sort by index in descending order along rows:

basic_df.sort_index(axis=0, ascending=False)

Sort by index in descending order along columns:

basic_df.sort_index(axis=1, ascending=False)

Sort rows based on values in a specific column (e.g., "epsilon"):

mixed_df.sort_values(by="epsilon")

Note: axis=0 refers to row-wise operations (index axis), while axis=1 targets column-wise behavior (columns axis). Setting ascending=False reverses sort direction.