2. NumPy

NumPy provides efficient multidimensional arrays.

import numpy as np

x = np.array([1, 2, 3])

Important properties

x.shape
x.ndim
x.size
x.dtype

Matrix:

X = np.array([
    [1, 2, 3],
    [4, 5, 6]
])

Shape:

(2, 3)

2 rows
3 columns

Vectorized operations

x * 2
x + 5
x.mean()
x.sum()

Prefer vectorized operations over manually looping through every value when possible.

ML convention

X.shape = (samples, features)

Example:

(1000, 10)

1000 samples
10 features each

3. Pandas

Pandas is useful for working with tabular data.

import pandas as pd

df = pd.read_csv("customers.csv")

Inspect data

df.head()
df.tail()
df.shape
df.columns
df.dtypes
df.info()
df.describe()

Missing values:

df.isna().sum()

Select column:

df["age"]

Multiple columns:

df[["age", "income"]]

Filter:

df[df["age"] >= 18]

loc vs iloc

df.loc[...]   # label-based
df.iloc[...]  # position-based

DataFrame vs Series

DataFrame → table / multiple columns
Series    → usually one column

ML convention

X = df.drop(columns=["target"])
y = df["target"]

X = input features
y = target/answer

4. Jupyter Notebooks

Notebook files:

.ipynb

They contain: - code cells - Markdown cells - outputs

Good for: - exploration - visualization - experiments - learning - EDA

Use normal .py modules for reusable/production application code.

Important: Notebook cells share state. Running cells out of order can create confusing results. Restart + Run All is a useful sanity check.