Introduction to Pandas

Introduction to Pandas

Pandas is the most powerful library for tabular data analysis in Python. Every AI project starts with loading and cleaning data.

The Two Core Objects

| Object | Description | When to use | |--------|------------|-------------| | Series | Single column with index | One-dimensional data | | DataFrame | 2D table | Any dataset |

import pandas as pd

scores = pd.Series([85, 92, 78, 96, 88],
                   index=["Ahmed", "Sara", "Ali", "Nour", "Omar"])
print(scores.mean())  # 87.8

data = {"name": ["Ahmed", "Sara"], "score": [85, 92]}
df = pd.DataFrame(data)

Loading Data

df = pd.read_csv("data.csv")
df = pd.read_json("data.json")
df = pd.read_excel("report.xlsx")

Initial Exploration

print(df.head(3))       # first 3 rows
print(df.shape)         # (num_rows, num_cols)
print(df.dtypes)        # each column type
print(df.info())        # comprehensive summary
print(df.describe())    # statistics for numeric columns

Selection and Filtering

ages    = df["age"]
subset  = df[["name", "salary"]]
ai_team = df[df["department"] == "AI"]
senior  = df[df["age"] > 30]
top     = df[(df["salary"] > 10000) & (df["dept"] == "ML")]
result  = df.query("age > 28 and salary > 9000")

Adding and Modifying Columns

df["monthly"] = df["salary"] / 12
df["level"] = df["age"].apply(lambda a: "senior" if a > 30 else "junior")
df.rename(columns={"salary": "annual"}, inplace=True)
df.drop(columns=["monthly"], inplace=True)

groupby — The Most Powerful Tool

dept_avg = df.groupby("department")["salary"].mean()

summary = df.groupby("department").agg({
    "salary": ["mean", "max"],
    "age": "mean",
})

Handling Missing Values

print(df.isnull().sum())
df_clean = df.dropna()
df["salary"].fillna(df["salary"].median(), inplace=True)

Lesson Summary

• DataFrame = flexible and powerful data table
• read_csv() = starting point for every AI project
• head/info/describe() = mandatory initial exploration
• Boolean indexing = easy data filtering
• groupby() = group and summarize data